2011-03-03 Sheriff Bot <webkit.review.bot@gmail.com>
[WebKit-https.git] / Tools / ChangeLog
1 2011-03-03  Sheriff Bot  <webkit.review.bot@gmail.com>
2
3         Unreviewed, rolling out r80237.
4         http://trac.webkit.org/changeset/80237
5         https://bugs.webkit.org/show_bug.cgi?id=55714
6
7         Windows bots aren't ready to be core (Requested by aroben on
8         #webkit).
9
10         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
11         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
12
13 2011-03-03  Mihai Parparita  <mihaip@chromium.org>
14
15         Reviewed by Tony Chang.
16
17         NRWT: AttributeError: TestRunner2 instance has no attribute '_cancel_workers'
18         https://bugs.webkit.org/show_bug.cgi?id=55694
19         
20         Fixes for interruptions in NRWT:
21         - Make TestRunInterruptedException be pickleable correctly (the base
22           Exception class defines a __reduce__ that does not include the reason)
23         - Fix ordering of returned arguments from TestRunner2._run_tests
24           (interrupted and keyboard_interrupted were reversed)
25         - Fix cancel_workers callsites (was using old name).
26         - In handle_exception re-raise actual exception instance that was thrown
27
28         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
29         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
30         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
31         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
32
33 2011-03-03  Qi Zhang  <qi.2.zhang@nokia.com>
34
35         Unreviewed.
36
37         Adding myself as a committer.
38
39         * Scripts/webkitpy/common/config/committers.py:
40
41 2011-03-03  Tony Gentilcore  <tonyg@chromium.org>
42
43         Reviewed by Mihai Parparita.
44
45         Teach sheriffbot to answer whois commands
46         https://bugs.webkit.org/show_bug.cgi?id=55687
47
48         * Scripts/webkitpy/tool/bot/irc_command.py:
49         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
50         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
51
52 2011-03-03  Victoria Kirst  <vrk@google.com>
53
54         Reviewed by Eric Carlson.
55
56         Enable media tests in chromium-gpu/test_expectations
57         https://bugs.webkit.org/show_bug.cgi?id=53608
58
59         Add media folder to the list of GPU test paths.
60
61         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
62
63 2011-03-03  Adam Roben  <aroben@apple.com>
64
65         Work around a bug in KURL's parsing of Windows-style absolute file: URLs
66
67         Fixes <http://webkit.org/b/55674> Many tests fail in WebKit2 mode on Windows due to
68         incorrect parsing of absolute Windows-style file: URLs
69
70         Reviewed by Tim Hatcher.
71
72         * WebKitTestRunner/TestInvocation.cpp:
73         (WTR::createWKURL): Add a "localhost" host to file: URLs we create to work around
74         <http://webkit.org/b/55683>.
75
76 2011-03-03  Tony Chang  <tony@chromium.org>
77
78         Reviewed by Dimitri Glazkov.
79
80         [chromium] Run kill-old-processes on the chromium-win bot too
81         https://bugs.webkit.org/show_bug.cgi?id=55630
82
83         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
84         * BuildSlaveSupport/win/kill-old-processes: Add wdiff.exe to the kill list.
85
86 2011-03-03  Carlos Garcia Campos  <cgarcia@igalia.com>
87
88         Reviewed by Martin Robinson.
89
90         [GTK] DRT needs implementation of EventSender.scheduleAsynchronousClick
91         https://bugs.webkit.org/show_bug.cgi?id=53960
92
93         * DumpRenderTree/gtk/EventSender.cpp:
94         (sendClick):
95         (scheduleAsynchronousClickCallback):
96
97 2011-03-03  Adam Roben  <aroben@apple.com>
98
99         Look for Windows-style paths when checking whether a test is an Inspector test
100
101         WTR part of <http://webkit.org/b/55672> <rdar://problem/9080867> All inspector tests time
102         out or crash on Windows 7 Release (WebKit2 Tests)
103
104         Reviewed by Darin Adler.
105
106         * WebKitTestRunner/TestInvocation.cpp:
107         (WTR::shouldOpenWebInspector): Also look for backslashes, since that's what we'll get on
108         Windows.
109
110 2011-03-03  Adam Roben  <aroben@apple.com>
111
112         Make Windows 7 Release (Tests) a core builder
113
114         We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
115         give it a few more days of greenness first.
116
117         Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
118
119         Reviewed by Andreas Kling.
120
121         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
122         (BuildBot): Added a regexp to match Windows 7 Release (Tests).
123
124         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
125         (BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
126         Updated the core regexps to match the buildbot module. Updated the expected results to match
127         the current set of core builders.
128
129 2011-03-03  Benjamin Poulain  <benjamin.poulain@nokia.com>
130
131         Reviewed by Andreas Kling.
132
133         Adding myself as a reviewer.
134
135         * Scripts/webkitpy/common/config/committers.py:
136
137 2011-03-03  Lukasz Slachciak  <l.slachciak@samsung.com>
138
139         Reviewed by Xan Lopez.
140
141         [GTK] Extended application cache database API and added unit tests file.
142         https://bugs.webkit.org/show_bug.cgi?id=55335
143
144         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: removed external declaration
145         of webkit_application_cache_set_maximum_size due to API exposal.
146
147 2011-03-02  Ojan Vafai  <ojan@chromium.org>
148
149         Reviewed by Adam Barth.
150
151         make webkit-patch upload work for security bugs
152         https://bugs.webkit.org/show_bug.cgi?id=55655
153
154         The problem is that we shell out to prepare-ChangeLog,
155         which does not authenticate with bugzilla. All it needs from
156         bugzilla is the bug title though. We just pass it through from
157         webkit-patch instead.
158
159         * Scripts/prepare-ChangeLog:
160         * Scripts/webkitpy/common/net/bugzilla/bug.py:
161         * Scripts/webkitpy/tool/steps/preparechangelog.py:
162
163 2011-03-01  Ojan Vafai  <ojan@chromium.org>
164
165         Reviewed by Tony Chang.
166
167         [test-dashboard] make gtest JSON match layout-tests JSON
168         https://bugs.webkit.org/show_bug.cgi?id=55556
169
170         http://trac.webkit.org/changeset/80090 stopped converting time_ms to t.
171         Make this code match.
172
173         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
174
175 2011-03-02  Dirk Pranke  <dpranke@chromium.org>
176
177         Reviewed by Tony Chang.
178
179         NRWT - implement Linux Hardy 64-bit port support. This involves
180         adding support for non-'x86' architectures and updating the
181         "deduplicate_tests" script to correctly understand that
182         port.name() is not always equal to basename(port.baseline_path()).
183
184         This change introduces two new port names: 'chromium-linux-x86'
185         and 'chromium-linux-x86_64'. Using the prior 'chromium-linux'
186         flag will cause the code to determine which one to use at
187         runtime, just like we currently do with 'chromium-win'.
188
189         https://bugs.webkit.org/show_bug.cgi?id=55535
190
191         * Scripts/webkitpy/common/system/executive_mock.py:
192           Handle 'return_stderr' keyword arg to run_command()
193         * Scripts/webkitpy/layout_tests/deduplicate_tests.py:
194         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
195         * Scripts/webkitpy/layout_tests/port/base.py:
196         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
197         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
198         * Scripts/webkitpy/layout_tests/port/factory.py:
199         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
200
201 2011-03-02  Kevin Ollivier  <kevino@theolliviers.com>
202
203         [wx] Build fix, make sure we put the wxWebKit dylib in the right directory and
204         update symlinks accordingly.
205
206         * wx/packaging/build-mac-installer.py:
207
208 2011-03-02  Sheriff Bot  <webkit.review.bot@gmail.com>
209
210         Unreviewed, rolling out r80139.
211         http://trac.webkit.org/changeset/80139
212         https://bugs.webkit.org/show_bug.cgi?id=55620
213
214         failing and crashing tests on Snow Leopard bot (Requested by
215         mihaip1 on #webkit).
216
217         * DumpRenderTree/LayoutTestController.cpp:
218         (LayoutTestController::staticFunctions):
219         * DumpRenderTree/LayoutTestController.h:
220         * DumpRenderTree/chromium/LayoutTestController.cpp:
221         (LayoutTestController::LayoutTestController):
222         * DumpRenderTree/chromium/LayoutTestController.h:
223         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
224         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
225         * DumpRenderTree/mac/UIDelegate.h:
226         * DumpRenderTree/mac/UIDelegate.mm:
227         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
228         * DumpRenderTree/qt/LayoutTestControllerQt.h:
229         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
230         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
231
232 2011-03-02  Adam Roben  <aroben@apple.com>
233
234         Wait for force a paint in WebKitTestRunner until we're actually ready to dump the output
235
236         This matches DumpRenderTree on Windows.
237
238         Fixes <http://webkit.org/b/55469> <rdar://problem/9068539> REGRESSION (r79863): Lots of
239         dom/html/level2/html/HTMLFrameElement*.html tests crashing in FrameView::paintOverhangAreas
240         in WebKit2
241
242         Reviewed by Beth Dakin.
243
244         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
245         (WTR::InjectedBundlePage::dump): Moved code to force a paint here...
246         (WTR::InjectedBundlePage::didFinishLoadForFrame): ...from here.
247
248 2011-03-02  Daniel Cheng  <dcheng@chromium.org>
249
250         Reviewed by David Levin.
251
252         Add feature define for data transfer items
253         https://bugs.webkit.org/show_bug.cgi?id=55510
254
255         * Scripts/build-webkit:
256
257 2011-03-02  Adam Roben  <aroben@apple.com>
258
259         Look in the new-to-SnowLeopard location for crash logs on SnowLeopard and newer
260
261         ReportCrash saves logs to ~/Library/Logs/CrashReporter on Leopard, but to
262         ~/Library/Logs/DiagnosticReports on SnowLeopard. old-run-webkit-tests was only looking in
263         the former location, but was getting lucky on some SnowLeopard machines because symlinks
264         were being created there pointing into DiagnosticReports. For machines without these
265         symlinks, crash logs were not getting captured.
266
267         Fixes <http://webkit.org/b/55607> old-run-webkit-tests doesn't capture crash logs on some
268         SnowLeopard machines
269
270         Reviewed by Eric Seidel.
271
272         * Scripts/old-run-webkit-tests:
273         (captureSavedCrashLog): Look in ~/Library/Logs/DiagnosticReports on SnowLeopard and newer.
274
275 2011-03-02  Adam Roben  <aroben@apple.com>
276
277         Don't try to capture crash logs on platforms that haven't implemented that feature
278
279         Fixes <http://webkit.org/b/55504> Use of uninitialized value in numeric lt (<) at
280         Tools/Scripts/old-run-webkit-tests line 1778 seen on Qt bots after a crash
281
282         Reviewed by Joseph Pecoraro.
283
284         * Scripts/old-run-webkit-tests:
285         (captureSavedCrashLog): Bail if we don't have a crash log glob.
286         (findNewestFileMatchingGlob): Fix a buggy test for when no paths match the glob. This isn't
287         strictly required to fix bug 55504, but seems worthwhile.
288
289 2011-01-26  John Knottenbelt  <jknotten@chromium.org>
290
291         Reviewed by Dmitry Titov.
292
293         Detach Geolocation from Frame when Page destroyed.
294         https://bugs.webkit.org/show_bug.cgi?id=52877
295
296         Extend the layout test controller to expose the number of pending
297         geolocation requests, so that we can test that the requests have
298         been cancelled on page close.
299
300         * DumpRenderTree/LayoutTestController.cpp:
301         (numberOfPendingGeolocationPermissionRequestsCallback):
302         (LayoutTestController::staticFunctions):
303         * DumpRenderTree/LayoutTestController.h:
304         * DumpRenderTree/chromium/LayoutTestController.cpp:
305         (LayoutTestController::LayoutTestController):
306         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
307         * DumpRenderTree/chromium/LayoutTestController.h:
308         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
309         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
310         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
311         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
312         * DumpRenderTree/mac/UIDelegate.h:
313         * DumpRenderTree/mac/UIDelegate.mm:
314         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
315         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
316         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
317         * DumpRenderTree/qt/LayoutTestControllerQt.h:
318         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
319         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
320         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
321         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
322
323
324 2011-03-01  Kent Tamura  <tkent@chromium.org>
325
326         Reviewed by Dimitri Glazkov.
327
328         Assertion fails when a form validation bubble appears
329         https://bugs.webkit.org/show_bug.cgi?id=55550
330
331         Test: fast/forms/interactive-validation-attach-assertion.html
332
333         * DumpRenderTree/chromium/WebPreferences.cpp:
334         (WebPreferences::applyTo): Enable the interactive validation feature by
335           default. This matches to Chromium browser.
336
337 2011-03-01  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
338
339         Reviewed by Kent Tamura.
340
341         Remove efl port style exception of Webcore/JavaScriptCore
342         https://bugs.webkit.org/show_bug.cgi?id=55549
343
344         WebCore and JavaScriptCore need to adhere WebKit coding style.
345
346         * Scripts/webkitpy/style/checker.py:
347
348 2011-03-01  Tony Chang  <tony@chromium.org>
349
350         Reviewed by Ojan Vafai.
351
352         [chromium] clean up temp files on chromium-mac builder too
353         https://bugs.webkit.org/show_bug.cgi?id=55525
354
355         * BuildSlaveSupport/chromium/remove-crash-logs:
356
357 2011-02-28  Ojan Vafai  <ojan@chromium.org>
358
359         Reviewed by Tony Chang.
360
361         wrap json in a function call to afford cross-domain loading
362         https://bugs.webkit.org/show_bug.cgi?id=55353
363
364         Also delete the code that compacts the JSON. It turns out this
365         breaks the rebaseline tool and is probably a premature optimization anyways.
366
367         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
368         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
369         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
370         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
371
372 2011-03-01  Sheriff Bot  <webkit.review.bot@gmail.com>
373
374         Unreviewed, rolling out r80079.
375         http://trac.webkit.org/changeset/80079
376         https://bugs.webkit.org/show_bug.cgi?id=55547
377
378         "Broke the Win debug build?" (Requested by dcheng on #webkit).
379
380         * Scripts/build-webkit:
381
382 2011-03-01  Mark Rowe  <mrowe@apple.com>
383
384         Reviewed by Sam Weinig.
385
386         Update verification scripts to handle WebKit2.
387
388         * Scripts/check-for-global-initializers: Skip object files that use RefCountedLeakChecker.
389         * Scripts/check-for-inappropriate-files-in-framework: Add some inappropriate file types for
390         WebKit2.
391
392 2011-03-01  Daniel Cheng  <dcheng@chromium.org>
393
394         Reviewed by David Levin.
395
396         Add feature define for data transfer items
397         https://bugs.webkit.org/show_bug.cgi?id=55510
398
399         * Scripts/build-webkit:
400
401 2011-02-23  Joseph Pecoraro  <joepeck@webkit.org>
402
403         Reviewed by Timothy Hatcher.
404
405         All Console Messages should be passed to ChromeClients.
406         https://bugs.webkit.org/show_bug.cgi?id=54926
407
408         Update DumpRenderTree's UIDelegate to use the new console message delegate.
409
410         * DumpRenderTree/mac/UIDelegate.mm:
411         (-[UIDelegate webView:addMessageToConsole:withSource:]):
412
413 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
414
415         Reviewed by Adam Barth.
416
417         Tragically, we stopped running the python unit tests
418         on Chromium Win just as I was getting them to be green.
419         This patch turns them back on.
420
421         https://bugs.webkit.org/show_bug.cgi?id=55521
422
423         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
424
425 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
426
427         Reviewed by Adam Roben.
428
429         Disable
430         webkitpy.common.system.executive_unittest.ExecutiveTest.\
431            test_kill_all fails on win until it can be properly triaged.
432
433         https://bugs.webkit.org/show_bug.cgi?id=54790
434
435         * Scripts/webkitpy/common/system/executive_unittest.py:
436
437     2011-03-01  Helder Correia  <helder@sencha.com>
438
439             Unreviewed. Add myself to committers.py.
440
441             * Scripts/webkitpy/common/config/committers.py:
442
443 2011-03-01  Adam Roben  <aroben@apple.com>
444
445         Give up if a crash log for the web process is taking too long to be saved on Windows
446
447         This should keep the bots from getting stuck while waiting for a crash log, if writing a
448         crash log ever hangs or the UI process for some reason doesn't notice it's finished.
449
450         Fixes <http://webkit.org/b/55499> WebKitTestRunner can hang forever waiting for a crash log
451         to be saved for the web process
452
453         Reviewed by Steve Falkenburg.
454
455         * WebKitTestRunner/win/TestControllerWin.cpp:
456         (WTR::runRunLoopUntil): Moved code to run the run loop here here from platformRunUntil.
457         Generalized the code slightly to handle the optional object to wait on.
458         (WTR::TestController::platformRunUntil): Use the new runRunLoopUntil function, and also use
459         it (with a timeout) when waiting for a crash log for the web process to be saved. This will
460         prevent us from waiting forever.
461
462 2011-03-01  Adam Roben  <aroben@apple.com>
463
464         Give WebKitLauncherWin's PDB file a name different from WebKit's
465
466         WebKitLauncherWin (which builds WebKit.exe) and WebKit (which builds WebKit.dll) were
467         fighting over the same PDB file, meaning that one of them was going to lose and not have
468         symbols. WebKitLauncherWin's PDB file is now named WebKitLauncherWin.pdb.
469
470         Fixes <http://webkit.org/b/55495> Crashes in WebKit.dll have bad backtraces for some test
471         runs (due to missing symbols)
472
473         Reviewed by David Kilzer.
474
475         * WebKitLauncherWin/WebKitLauncherWinCommon.vsprops: Use the project name,
476         "WebKitLauncherWin", instead of the target name, "WebKit", when naming the PDB file.
477
478 2011-03-01  Adam Roben  <aroben@apple.com>
479
480         Count web process crashes toward the --exit-after-n-crashes-or-timeouts limit
481
482         Fixes <http://webkit.org/b/55492> Web process crashes aren't counted toward the
483         --exit-after-n-crashes-or-timeouts limit
484
485         Reviewed by David Kilzer.
486
487         * Scripts/old-run-webkit-tests:
488         (stopRunningTestsEarlyIfNeeded): Include web process crashes in the computation, and print
489         the number of tests that caused the web process to crash when exiting early.
490
491 2011-03-01  Adam Roben  <aroben@apple.com>
492
493         Save a crash log when the web process crashes
494
495         On Windows, WebKitTestRunner now detects when the web process is crashing and waits to exit
496         until it has finished crashing, which guarantees that the crash log will have had time to be
497         saved, too. On Mac, we always wait until ReportCrash has exited before capturing the crash
498         log, so all we have to do is choose the right crash log out of the CrashReporter directory.
499
500         Fixes <http://webkit.org/b/44121> <rdar://problem/8320759> When the web process crashes and
501         a crash log is being saved, WebKitTestRunner thinks the web process has become unresponsive
502
503         Reviewed by Sam Weinig.
504
505         * Scripts/old-run-webkit-tests:
506         (testCrashedOrTimedOut): Don't kill WebKitTestRunner when the web process crashes. It will
507         kill itself. On Windows, this will cause us to wait until the crash log has been saved. On
508         Mac, it should have no effect. Capture saved crash logs for web process crashes, too.
509         (captureSavedCrashLog): Added $webProcessCrashed argument. On Mac, look for
510         WebProces_*.crash files when the web process crashes.
511
512         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
513         (WTR::InjectedBundle::initialize): Added an initializationUserData argument. Updated for
514         initializePlatformDefaults -> platformInitialize rename. Pass the initializationUserData
515         along to platformInitialize.
516
517         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: See above.
518
519         * WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
520         (WKBundleInitialize): Pass along the initializationUserData to the InjectedBundle.
521
522         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
523         (WTR::InjectedBundle::platformInitialize):
524         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
525         (WTR::InjectedBundle::platformInitialize):
526         Updated function signature.
527
528         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
529         (WTR::exceptionFilter): Added. Tells the UI process we're crashing by signaling the
530         webProcessCrashingEvent, then lets the crash continue as normal.
531
532         (WTR::InjectedBundle::platformInitialize): Hook up exceptionFilter. Retrieve the name of the
533         event we should use to tell the UI process we're crashing from the initializationUserData,
534         and get a handle to that event.
535
536         * WebKitTestRunner/TestController.cpp:
537         (WTR::TestController::TestController): Initialize new members.
538         (WTR::TestController::processDidCrash): Removed unnecessary WKPageRef argument. Changed to
539         only print the "#CRASHED - WebProcess" message once, since this can be called more than once
540         when a crash log is being saved on Windows. Exit right away if specified. (This is the
541         default.)
542
543         * WebKitTestRunner/TestController.h: Added new members.
544
545         * WebKitTestRunner/win/TestControllerWin.cpp:
546         (WTR::TestController::platformInitialize): Set up the event the web process will use to tell
547         us it's crashing.
548         (WTR::TestController::platformRunUntil): Pass MWMO_INPUTAVAILABLE to
549         ::MsgWaitForMultipleObjectsEx so we'll process messages that have already been seen by
550         ::PeekMessage. (This is unrelated to the bug fix.) Notice when the webProcessCrashingEvent
551         has been signaled. When this happens, print the "#CRASHED - WebProcess" message right away
552         so the test harness will know the web process has crashed and not try to kill us, then wait
553         for the web process to finish crashing so a crash log will have time to be saved.
554         (WTR::toWK): Simple hepler function.
555         (WTR::TestController::platformInitializeContext): Pass along the name of the event the web
556         process should use to tell us it is crashing in the context's initialization user data.
557
558 2011-03-01  Dimitri Glazkov  <dglazkov@chromium.org>
559
560         Reviewed by Tony Gentilcore.
561
562         Move the checks to exclude Python/Perl tests for Chromium Win build to the right place.
563         https://bugs.webkit.org/show_bug.cgi?id=55476
564
565         Chromium Win uses BuildAndTestFactory, not TestFactory.
566
567         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Moved to the right place.
568
569 2011-03-01  Adam Roben  <aroben@apple.com>
570
571         Test that the WebKit2 responsiveness timer doesn't fire too early
572
573         Test for <http://webkit.org/b/55417> <rdar://problem/9065287> RunLoop::Timer fires 1000x too
574         early on Windows
575
576         Reviewed by Anders Carlsson.
577
578         * TestWebKitAPI/PlatformUtilities.h: Added sleep().
579
580         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp: Added.
581         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
582         (TestWebKitAPI::didFinishLoadForFrame):
583         (TestWebKitAPI::processDidBecomeUnresponsive):
584         (TestWebKitAPI::setInjectedBundleClient):
585         (TestWebKitAPI::setPageLoaderClient):
586         Simple helper functions.
587
588         (TestWebKitAPI::TEST): Load an HTML file to make sure the web process is initialized. Then
589         tell the web process to pause and press the spacebar key. The spacebar keypress should cause
590         the responsiveness timer to start, but the web process should unpause before it has a chance
591         to fire. Run until the web process has unpaused, and assert that the timer didn't fire.
592
593         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly_Bundle.cpp: Added.
594         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::ResponsivenessTimerDoesntFireEarlyTest):
595         Call up to the base class.
596         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::didReceiveMessage): When asked to
597         pause, sleep for 0.5 seconds, then send back a message saying we paused.
598
599         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
600         (TestWebKitAPI::Util::sleep): Added. Calls through to usleep.
601         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
602         (TestWebKitAPI::Util::sleep): Added. Calls through to ::Sleep.
603
604         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
605         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
606         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
607         Added new files.
608
609 2011-03-01  Andras Becsi  <abecsi@webkit.org>
610
611         Reviewed by Csaba Osztrogonác.
612
613         [Qt] Clean up the project files and move common options to WebKit.pri.
614
615         * WebKitTestRunner/qt/WebKitTestRunner.pro: Deduplicate.
616
617
618 2011-03-01  Adam Roben  <aroben@apple.com>
619
620         Skip another multiprocessing test on Windows
621
622         It is sometimes failing (and may be leaving child Python processes around).
623         <http://webkit.org/b/55087> tracks the failure.
624
625         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
626         (FuncitonTests.test_get__processes): Skip this test on Windows.
627
628 2011-03-01  Ilya Tikhonovsky  <loislo@chromium.org>
629
630         Reviewed by Pavel Feldman.
631
632         Web Inspector: reduce inspector tests flakiness.
633         https://bugs.webkit.org/show_bug.cgi?id=55401
634
635         * DumpRenderTree/chromium/TestShell.cpp:
636         (TestShell::runFileTest):
637         * DumpRenderTree/chromium/WebPreferences.cpp:
638         (WebPreferences::reset):
639
640 2011-02-28  Mihai Parparita  <mihaip@chromium.org>
641
642         Reviewed by James Robinson.
643
644         NRWT should still upload results when it exits early due to too many failures
645         https://bugs.webkit.org/show_bug.cgi?id=55432
646         
647         Only keyboard interrupts should prevent results from being uploaded.
648
649         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
650         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
651
652 2011-02-28  Alice Liu  <alice.liu@apple.com>
653
654         Reviewed by Adam Roben and Gavin Barraclough.
655
656         Test for https://bugs.webkit.org/show_bug.cgi?id=54898
657         InjectedBundleNodeHandle dies too early in WKBundleHitTestResultGetNodeHandle.
658
659         * TestWebKitAPI/InjectedBundleController.h:
660         (TestWebKitAPI::InjectedBundleController::bundle):
661         * TestWebKitAPI/PlatformWebView.h:
662         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp: Added.
663         (TestWebKitAPI::didFinishLoadForFrame):
664         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
665         (TestWebKitAPI::setPageLoaderClient):
666         (TestWebKitAPI::TEST):
667         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp: Added.
668         (TestWebKitAPI::HitTestResultNodeHandleTest::HitTestResultNodeHandleTest):
669         (TestWebKitAPI::HitTestResultNodeHandleTest::getContextMenuFromDefaultMenu):
670         (TestWebKitAPI::HitTestResultNodeHandleTest::didCreatePage):
671         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
672         (TestWebKitAPI::PlatformWebView::simulateRightClick):
673         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
674         (TestWebKitAPI::PlatformWebView::simulateRightClick):
675         
676         Added files to:
677         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
678         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
679         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
680
681 2011-02-28  Adam Roben  <aroben@apple.com>
682
683         Start running InjectedBundle API tests on Windows
684
685         A new project, TestWebKitAPIInjectedBundle, builds the injected bundle DLL.
686
687         Fixes <http://webkit.org/b/55420> run-api-tests doesn't run injected bundle tests on Windows
688
689         Reviewed by Darin Adler.
690
691         * TestWebKitAPI/Configurations/TestWebKitAPIDebug.vsprops: Added.
692         * TestWebKitAPI/Configurations/TestWebKitAPIDebugAll.vsprops: Added.
693         * TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops: Added.
694         * TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops: Added.
695         * TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops: Added.
696         These are just extracted from TestWebKitAPI.vcproj.
697
698         * TestWebKitAPI/Configurations/TestWebKitAPIInjectedBundleCommon.vsprops: Added. Just sets
699         up the DLL's filename.
700
701         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Fixed DLL name.
702         (TestWebKitAPI::Util::createInjectedBundlePath): Changed to use kCFAllocatorNull instead of
703         0 (== kCFAllocatorDefault) so that CF won't try to deallocate our C string.
704
705         * TestWebKitAPI/win/TestWebKitAPI.sln: Added TestWebKitAPIInjectedBundle. Removed the bogus
706         "all" configuration.
707
708         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Changed to use the new .vsprops files. Added
709         injected bundle tests. Let VS resort things.
710
711         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Copied from Tools/TestWebKitAPI/win/TestWebKitAPI.vcproj.
712
713 2011-02-28  Adam Roben  <aroben@apple.com>
714
715         Add a test for the paint rect passed via WM_PAINT to windowless plugins
716
717         Test for <http://webkit.org/b/55365> <rdar://problem/9031089> REGRESSION (r79040): WebKit2:
718         Large portions of pages with plugins paint black
719
720         Reviewed by Anders Carlsson.
721
722         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
723         (PluginTest::NPP_HandleEvent):
724         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
725         Added NPP_HandleEvent. Just returns 0 at this level.
726
727         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowlessPaintRectCoordinates.cpp: Added.
728         (WindowlessPaintRectCoordinates::WindowlessPaintRectCoordinates): Call up to the base class.
729         (WindowlessPaintRectCoordinates::NPP_New): Mark ourselves as windowless.
730         (WindowlessPaintRectCoordinates::NPP_HandleEvent): Check that the paint rect passed via
731         WM_PAINT has the expected coordinates.
732
733         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
734         (NPP_HandleEvent): Give the PluginTest a chance to handle the event before doing anything
735         else.
736
737         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new file, let VS
738         resort things.
739
740         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
741         (WTR::InjectedBundlePage::didFinishLoadForFrame): Added a call to WKBundlePageForceRepaint.
742         This is roughly equivalent to the call to -[WebView displayIfNeeded] in -[FrameLoadDelegate
743         webView:didFinishLoadForFrame:] in DumpRenderTree.
744
745         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
746         (WTR::PlatformWebView::PlatformWebView): Tell the WKView it's in a window so that plugins
747         can start running.
748
749 2011-02-28  Adam Roben  <aroben@apple.com>
750
751         Tell MiniBrowser's WKViews that they're in a window
752
753         Fixes <http://webkit.org/b/55364> Plugins don't work in MiniBrowser on Windows
754
755         Reviewed by Anders Carlsson.
756
757         * MiniBrowser/win/BrowserView.cpp:
758         (BrowserView::create): Call WKViewSetIsInWindow after creating the view.
759
760 2011-02-28  Alexis Menard  <alexis.menard@openbossa.org>
761
762         Reviewed by Csaba Osztrogonác.
763
764         The new QML tests require QML_IMPORT_PATH variable to be set in order to
765         load the proper plugin.
766
767         * Scripts/run-qtwebkit-tests:
768
769 2011-02-28  Joone Hur  <joone.hur@collabora.co.uk>
770
771         Unreviewed, fix spelling mistake.
772
773         hardware-accelated => hardware-accelerated.
774
775         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
776
777 2011-02-28  Kristian Amlie  <kristian.amlie@nokia.com>
778
779         Reviewed by Andreas Kling.
780
781         Made sure that the build-webkit --qmake argument is always respected.
782
783         Previously, it would be respected during the build, but not during
784         the early feature detection. The build would also fail if qmake was
785         not in your path, even if you specified --qmake with a valid qmake
786         binary.
787
788         [Qt] WebKit patches required to work with a modularized version of Qt
789         https://bugs.webkit.org/show_bug.cgi?id=53916
790
791         * Scripts/build-webkit:
792         * Scripts/webkitdirs.pm:
793
794 2011-02-27  Ojan Vafai  <ojan@chromium.org>
795
796         Fix failing tests from http://trac.webkit.org/changeset/79837.
797
798         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
799         This was just a mistake that would break if someone tried printing with this configuration.
800         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
801         Method signature changed. Didn't notice that this method was being calling in a unittest.
802
803 2011-02-24  Hayato Ito  <hayato@chromium.org>
804
805         Reviewed by Eric Seidel.
806
807         [NRWT] Remove test_types/* classes.
808
809         This is the last patch in the series of efforts which tried to remove test_types classes.
810         A single_test_runner module will take over responsibilities of test_type classes.
811
812         From this patch, we won't display statistics per test_type even if a '--verbose' option is given.
813
814         https://bugs.webkit.org/show_bug.cgi?id=55123
815
816         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
817         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
818         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
819         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
820         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
821         * Scripts/webkitpy/layout_tests/test_types/__init__.py: Removed.
822         * Scripts/webkitpy/layout_tests/test_types/image_diff.py: Removed.
823         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py: Removed.
824         * Scripts/webkitpy/layout_tests/test_types/test_type_base_unittest.py: Removed.
825         * Scripts/webkitpy/layout_tests/test_types/text_diff.py: Removed.
826
827 2011-02-25  Ojan Vafai  <ojan@chromium.org>
828
829         Reviewed by Tony Chang.
830
831         Change results.json format to the one used by unexpected_results.json
832         https://bugs.webkit.org/show_bug.cgi?id=52267
833
834         Also add runtimes in milliseconds to the JSON and make the output format more compact.
835         Named the file full_results.json to avoid conflicting with the results.json
836         file the test-results server currently serves up.
837
838         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
839         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
840         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
841         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
842         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
843         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
844
845 2011-02-27  Adam Roben  <aroben@apple.com>
846
847         Followup fix to r79827
848
849         * DumpRenderTree/win/UIDelegate.cpp:
850         (UIDelegate::webViewAddMessageToConsole): Replaced call to urlSuitableForTestResult with
851         lastPathComponent. This matches what we did here before r79827, and matches Mac.
852
853 2011-02-27  Dan Bernstein  <mitz@apple.com>
854
855         Reviewed by Adam Roben.
856
857         Make bisect-builds print a trac link for the regression range
858         https://bugs.webkit.org/show_bug.cgi?id=55331
859
860         * Scripts/bisect-builds:
861
862 2011-02-27  Adam Roben  <aroben@apple.com>
863
864         Use iswalpha instead of isalpha when dealing with wchar_ts in EditingDelegate
865
866         Fixes <http://webkit.org/b/55062> <rdar://problem/9059907> Crash beneath
867         EditingDelegate::checkSpellingOfString when running fast/forms/input-text-maxlength.html or
868         fast/forms/input-text-paste-maxlength.html on Windows with full page heap enabled
869
870         Reviewed by Anders Carlsson.
871
872         * DumpRenderTree/win/EditingDelegate.cpp:
873         (indexOfFirstWordCharacter):
874         (wordLength):
875         Use iswalpha instead of isalpha, since TCHAR is really just wchar_t.
876
877 2011-02-27  Adam Roben  <aroben@apple.com>
878
879         Make ResourceLoadDelegate print URLs relative to the main resource on Windows
880
881         This matches Mac. (In fact, the code was ported from Mac!)
882
883         Fixes <http://webkit.org/b/55328> URLs printed by ResourceLoadDelegate on Windows don't
884         match Mac
885
886         Reviewed by Anders Carlsson.
887
888         * DumpRenderTree/win/DumpRenderTree.cpp:
889         (substringFromIndex): Added. Emulates -[NSString substringFromIndex:].
890         (urlSuitableForTestResult): Ported code from -[NSURL(DRTExtras)
891         _drt_descriptionSuitableForTestResult].
892         (cfStringRefToWString): Moved here from LayoutTestControllerWin.cpp.
893
894         * DumpRenderTree/win/DumpRenderTreeWin.h: Added declaration of cfStringRefToWString.
895
896         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Moved cfStringRefToWString to
897         DumpRenderTree.cpp.
898
899 2011-02-27  Adam Roben  <aroben@apple.com>
900
901         Change the order of Chang Shu's email addresses so the first one is their Bugzilla username
902
903         This makes Bugzilla autocompletion work correctly.
904
905         * Scripts/webkitpy/common/config/committers.py:
906
907 2011-02-27  Patrick Gansterer  <paroga@webkit.org>
908
909         Reviewed by Andreas Kling.
910
911         [Qt] fast/loader/user-stylesheet-fast-path.html fails
912         https://bugs.webkit.org/show_bug.cgi?id=50911
913
914         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
915         (LayoutTestController::setUserStyleSheetLocation):
916         Use QUrl::fromEncoded since the passed string is already encoded.
917
918 2011-02-27  Adam Roben  <aroben@apple.com>
919
920         Ensure $testResultsDirectory is an absolute path before setting up Windows crash log saving
921
922         Windows 7 (and maybe Vista?) launches the post-mortem debugger with a working directory that
923         is different from old-run-webkit-test's. Because of this, the path to the crash log file
924         (which is relative to $testResultsDirectory) must be an absolute path. See the bug for more
925         details.
926
927         Fixes <http://webkit.org/b/55318> Crash logs aren't saved on Windows 7 when a relative
928         --results-directory path is used (like on the test slaves)
929
930         Reviewed by Dan Bernstein.
931
932         * Scripts/old-run-webkit-tests: Moved the call to setUpWindowsCrashLogSaving after we've
933         made $testResultsDirectory an absolute path.
934
935 2011-02-26  Dominic Mazzoni  <dmazzoni@google.com>
936
937         Reviewed by Adam Barth.
938
939         m_dumpAccessibilityNotifications should be initialized.
940         https://bugs.webkit.org/show_bug.cgi?id=54874
941
942         * DumpRenderTree/chromium/AccessibilityController.cpp:
943         (AccessibilityController::AccessibilityController):
944
945 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
946
947         Reviewed by Pavel Feldman.
948
949         DumpRenderTree should reset frame opener between tests.
950         https://bugs.webkit.org/show_bug.cgi?id=54874
951
952         Added clearOpener() calls between tests.
953         Some tests are using window.open() to change current window location.
954         This changes window.opener property whose nullity is checked before
955         showing xml viewer.
956
957         * DumpRenderTree/chromium/TestShell.cpp:
958         (TestShell::resetTestController):
959         * DumpRenderTree/gtk/DumpRenderTree.cpp:
960         (resetDefaultsToConsistentValues):
961         * DumpRenderTree/mac/DumpRenderTree.mm:
962         (resetWebViewToConsistentStateBeforeTesting):
963         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
964         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
965         * DumpRenderTree/win/DumpRenderTree.cpp:
966         (resetWebViewToConsistentStateBeforeTesting):
967         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
968         (WTR::InjectedBundlePage::reset):
969
970 2011-02-26  Yael Aharon  <yael.aharon@nokia.com>
971
972         Reviewed by Andreas Kling.
973
974         [Qt] Fix the focus in MiniBrowser.
975         https://bugs.webkit.org/show_bug.cgi?id=55288
976
977         Give focus to the web view when a load starts.
978
979         * MiniBrowser/qt/BrowserWindow.cpp:
980         (BrowserWindow::urlChanged):
981
982 2011-02-26  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
983
984         Reviewed by Adam Barth.
985
986         [EFL] Add coding style exceptions for EFL port
987         https://bugs.webkit.org/show_bug.cgi?id=54733
988
989         The EFL APIs use EFL naming style, which includes both lower-cased and camel-cased,
990         underscore-sparated values. But, style checking rule have found parameter_name errors
991         in efl directories so far.
992
993         * Scripts/webkitpy/style/checker.py: Add prameter_name rule's exception and efl directories.
994
995 2011-02-26  David Levin  <levin@chromium.org>
996
997         Reviewed by Darin Adler.
998
999         check-webkit-style should check for spaces after periods in comments
1000         https://bugs.webkit.org/show_bug.cgi?id=55253
1001
1002         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
1003         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests
1004           and fixed the name of one related test.
1005
1006 2011-02-26  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1007
1008         Reviewed by Andreas Kling.
1009
1010         Make it possible to test the targetdensity-dpi support
1011         https://bugs.webkit.org/show_bug.cgi?id=55142
1012
1013         Test the viewport meta tag feature targetdensity-dpi by
1014         adding extra arguments to dumpConfigurationForViewport
1015
1016         Add the new arguments to the shared header and all
1017         implementation.
1018
1019         * DumpRenderTree/LayoutTestController.cpp:
1020         (dumpConfigurationForViewportCallback):
1021         * DumpRenderTree/LayoutTestController.h:
1022         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1023         (LayoutTestController::dumpConfigurationForViewport):
1024         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1025         (LayoutTestController::dumpConfigurationForViewport):
1026         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1027         (LayoutTestController::dumpConfigurationForViewport):
1028         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1029         (LayoutTestController::dumpConfigurationForViewport):
1030         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1031         (LayoutTestController::dumpConfigurationForViewport):
1032         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1033
1034 2011-02-26  Sheriff Bot  <webkit.review.bot@gmail.com>
1035
1036         Unreviewed, rolling out r79764.
1037         http://trac.webkit.org/changeset/79764
1038         https://bugs.webkit.org/show_bug.cgi?id=55295
1039
1040         "broke Chromium builds" (Requested by rniwa on #webkit).
1041
1042         * DumpRenderTree/chromium/TestShell.cpp:
1043         (TestShell::resetTestController):
1044         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1045         (resetDefaultsToConsistentValues):
1046         * DumpRenderTree/mac/DumpRenderTree.mm:
1047         (resetWebViewToConsistentStateBeforeTesting):
1048         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1049         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1050         * DumpRenderTree/win/DumpRenderTree.cpp:
1051         (resetWebViewToConsistentStateBeforeTesting):
1052         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1053         (WTR::InjectedBundlePage::reset):
1054
1055 2011-02-26  Yongjun Zhang  <yongjun_zhang@apple.com>
1056
1057         Reviewed by David Kilzer.
1058
1059         https://bugs.webkit.org/show_bug.cgi?id=48781
1060
1061         Add a resource load delegate method to query if WebCore should paint the default broken image for failed images.
1062
1063         Add a new resource load client method (shouldPaintBrokenImage).  WebKit client can decide if WebCore
1064         should paint the default broken image when an image fails to load or decode.
1065
1066         * DumpRenderTree/LayoutTestController.cpp:
1067         (LayoutTestController::LayoutTestController):
1068         (setShouldPaintBrokenImageCallback):
1069         (LayoutTestController::staticFunctions):
1070         (LayoutTestController::setShouldPaintBrokenImage):
1071         * DumpRenderTree/LayoutTestController.h:
1072         (LayoutTestController::shouldPaintBrokenImage):
1073         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
1074         (-[ResourceLoadDelegate webView:shouldPaintBrokenImageForURL:]):
1075
1076 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
1077
1078         Reviewed by Pavel Feldman.
1079
1080         DumpRenderTree should reset frame opener between tests.
1081         https://bugs.webkit.org/show_bug.cgi?id=54874
1082
1083         Added clearOpener() calls between tests.
1084         Some tests are using window.open() to change current window location.
1085         This changes window.opener property whose nullity is checked before
1086         showing xml viewer.
1087
1088         * DumpRenderTree/chromium/TestShell.cpp:
1089         (TestShell::resetTestController):
1090         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1091         (resetDefaultsToConsistentValues):
1092         * DumpRenderTree/mac/DumpRenderTree.mm:
1093         (resetWebViewToConsistentStateBeforeTesting):
1094         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1095         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1096         * DumpRenderTree/win/DumpRenderTree.cpp:
1097         (resetWebViewToConsistentStateBeforeTesting):
1098         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1099         (WTR::InjectedBundlePage::reset):
1100
1101 2011-02-25  Lukasz Slachciak  <l.slachciak@samsung.com>
1102
1103         Reviewed by Kenneth Rohde Christiansen.
1104
1105         [EFL] Added EWebLauncher command line option to turn on/off frame flattening
1106         https://bugs.webkit.org/show_bug.cgi?id=54809
1107
1108         * EWebLauncher/main.c: Handling -f option added.
1109         (on_key_down):
1110         (browserCreate):
1111         (main):
1112
1113 2011-02-25  Sheriff Bot  <webkit.review.bot@gmail.com>
1114
1115         Unreviewed, rolling out r79324.
1116         http://trac.webkit.org/changeset/79324
1117         https://bugs.webkit.org/show_bug.cgi?id=55260
1118
1119         Leopard Debug is once again running out of address space while
1120         linking (Requested by aroben on #webkit).
1121
1122         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1123         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1124
1125 2011-02-25  Tony Chang  <tony@chromium.org>
1126
1127         Reviewed by James Robinson.
1128
1129         Clean up temp files left by crashing DRTs on chromium-win
1130         https://bugs.webkit.org/show_bug.cgi?id=55182
1131
1132         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1133         * BuildSlaveSupport/chromium/remove-crash-logs: Added.
1134
1135 2011-02-25  Adam Roben  <aroben@apple.com>
1136
1137         Double-check that modifying the registry worked
1138
1139         On Windows Vista/7 with UAC enabled, regtool will fail to modify the registry, but it will
1140         still return a successful exit code. So we double-check here that the value we tried to
1141         write to the registry was really written.
1142
1143         Fixes <http://webkit.org/b/55227> old-run-webkit-tests falsely claims to be saving crash
1144         logs on Windows Vista/7
1145
1146         Reviewed by Darin Adler.
1147
1148         * Scripts/old-run-webkit-tests:
1149         (writeRegistryString): After trying to write the string value, check that it was really
1150         written.
1151
1152 2011-02-25  Adam Roben  <aroben@apple.com>
1153
1154         Set the 32-bit post-mortem debugger on 64-bit OSes
1155
1156         The post-mortem debugger is controlled by some registry values in
1157         HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AeDebug. On 64-bit OSes, you can either
1158         set the values in that key, which will be used for 64-bit apps, or in the equivalent key
1159         under Wow6432Node, which will be used for 32-bit apps. Since DumpRenderTree and
1160         WebKitTestRunner are 32-bit, we want to use the latter location.
1161
1162         Fixes <http://webkit.org/b/55225> old-run-webkit-tests sets the post-mortem debugger for
1163         64-bit apps, but should set it for 32-bit apps instead
1164
1165         Reviewed by Darin Adler.
1166
1167         * Scripts/old-run-webkit-tests:
1168         (readRegistryString):
1169         (writeRegistryString):
1170         Pass --wow32 to regtool so it will get/set values beneath Wow6432Node, which is where
1171         Windows looks for the post-mortem debugger values when a 32-bit app crashes on a 64-bit OS.
1172         Passing --wow32 on a 32-bit OS will not have any effect.
1173
1174 2011-02-25  Adam Roben  <aroben@apple.com>
1175
1176         Clean up code to set up crash log saving on Windows
1177
1178         Fixes <http://webkit.org/b/55224> old-run-webkit-test's Windows registry code is ugly
1179
1180         Reviewed by David Kilzer.
1181
1182         * Scripts/old-run-webkit-tests:
1183         (readRegistryString):
1184         (writeRegistryString):
1185         Moved code to read strings from and write strings to the registry here...
1186
1187         (setUpWindowsCrashLogSaving): ...from here.
1188
1189 2011-02-25  Adam Roben  <aroben@apple.com>
1190
1191         Work around Cygwin's crash-suppression behavior
1192
1193         Cygwin calls ::SetErrorMode(SEM_FAILCRITICALERRORS), which any processes it launches will
1194         inherit. This is bad for testing/debugging, as it causes the post-mortem debugger not to be
1195         invoked. (Cygwin does this because it makes crashes more UNIX-y.) We reset the error mode
1196         when our test apps launch to work around Cygwin's behavior.
1197
1198         Fixes <http://webkit.org/b/55222> Test apps crash silently (without invoking post-mortem
1199         debugger) when launched from Cygwin 1.7
1200
1201         Reviewed by Darin Adler.
1202
1203         * DumpRenderTree/win/DumpRenderTree.cpp:
1204         (main):
1205         * TestWebKitAPI/win/main.cpp:
1206         (main):
1207         * WebKitAPITest/main.cpp:
1208         (main):
1209         * WebKitTestRunner/win/TestControllerWin.cpp:
1210         (WTR::TestController::platformInitialize):
1211         Call ::SetErrorMode(0) to undo Cygwin's folly.
1212
1213 2011-02-24  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
1214
1215         Reviewed by Andreas Kling.
1216
1217         [Qt] Revert the support for QNAM affined to a different thread.
1218         https://bugs.webkit.org/show_bug.cgi?id=55149
1219
1220         Qt 4.8 will have QNAM use its own thread internally by default,
1221         no need to keep this complexity in WebKit.
1222
1223         This mainly reverts:
1224         http://trac.webkit.org/changeset/73710
1225         http://trac.webkit.org/changeset/73712
1226
1227         * QtTestBrowser/launcherwindow.cpp:
1228         (LauncherWindow::~LauncherWindow):
1229         (LauncherWindow::initializeView):
1230         (LauncherWindow::createChrome):
1231         * QtTestBrowser/launcherwindow.h:
1232         (WindowOptions::WindowOptions):
1233         * QtTestBrowser/webpage.cpp:
1234         (WebPage::WebPage):
1235         * QtTestBrowser/webpage.h:
1236
1237 2011-02-24  Adam Barth  <abarth@webkit.org>
1238
1239         Reviewed by Eric Seidel.
1240
1241         Add a script for creating gypi files
1242         https://bugs.webkit.org/show_bug.cgi?id=55113
1243
1244         This is a trivial script to create gypi files from the source code in
1245         the current directory.
1246
1247         * Scripts/make-gypi: Added.
1248
1249 2011-02-24  Sheriff Bot  <webkit.review.bot@gmail.com>
1250
1251         Unreviewed, rolling out r79628.
1252         http://trac.webkit.org/changeset/79628
1253         https://bugs.webkit.org/show_bug.cgi?id=55195
1254
1255         "broke chromium win tests" (Requested by dpranke on #webkit).
1256
1257         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1258         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1259
1260 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
1261
1262         Reviewed by Mihai Parparita.
1263
1264         rebaseline-chromium-webkit-tests can't handle GPU variants
1265         https://bugs.webkit.org/show_bug.cgi?id=55002
1266
1267         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1268         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
1269
1270 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
1271
1272         Reviewed by Mihai Parparita.
1273
1274         nrwt: log command used to start web server to --verbose
1275
1276         https://bugs.webkit.org/show_bug.cgi?id=55188
1277
1278         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1279         * Scripts/webkitpy/layout_tests/port/http_server.py:
1280
1281 2011-02-24  Victoria Kirst  <vrk@google.com>
1282
1283         Reviewed by Mihai Parparita.
1284
1285         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
1286         https://bugs.webkit.org/show_bug.cgi?id=55089
1287
1288         This adds an alias into the media/ directory so that http/tests/media
1289         tests can access the media resources when running in an httpd process.
1290
1291         * Scripts/webkitperl/httpd.pm:
1292         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1293         * Scripts/webkitpy/layout_tests/port/http_server.py:
1294
1295 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
1296
1297         Reviewed by Mihai Parparita.
1298
1299         nrwt: enable multiple processes by default on Chromium Win.
1300
1301         https://bugs.webkit.org/show_bug.cgi?id=55163
1302
1303         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1304         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1305
1306 2011-02-24  Sam Weinig  <sam@webkit.org>
1307
1308         Attempt to fix Qt.
1309
1310         * WebKitTestRunner/TestInvocation.cpp:
1311
1312 2011-02-24  James Robinson  <jamesr@chromium.org>
1313
1314         Unreviewed, rolling out r79598.
1315         http://trac.webkit.org/changeset/79598
1316         https://bugs.webkit.org/show_bug.cgi?id=55089
1317
1318         Appears to be causing trouble with httpd on windows
1319
1320         * Scripts/webkitperl/httpd.pm:
1321         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1322         * Scripts/webkitpy/layout_tests/port/http_server.py:
1323
1324 2011-02-24  Sam Weinig  <sam@webkit.org>
1325
1326         Reviewed by Darin Adler.
1327
1328         Add basic pixel testing support to WebKitTestRunner
1329         https://bugs.webkit.org/show_bug.cgi?id=55161
1330
1331         * Scripts/old-run-webkit-tests:
1332         Don't print color space information when using WebKitTestRunner since
1333         we don't need to change the screen.
1334
1335         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1336         (WTR::InjectedBundle::initialize):
1337         Add call to initializePlatformDefaults.
1338         
1339         (WTR::InjectedBundle::done):
1340         Return a dictionary on completion with both the text output and pixel dump override.
1341
1342         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1343         Add declaration for initializePlatformDefaults.
1344
1345         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1346         (WTR::LayoutTestController::LayoutTestController):
1347         (WTR::LayoutTestController::dumpAsText):
1348         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1349         (WTR::LayoutTestController::shouldDumpPixels):
1350         Make setting dumpAsText override the dump pixels preference.
1351
1352         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: Added.
1353         (WTR::InjectedBundle::initializePlatformDefaults):
1354         Set platform defaults matching DumpRenderTree.
1355
1356         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp: Added.
1357         (WTR::InjectedBundle::initializePlatformDefaults):
1358         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp: Added.
1359         (WTR::InjectedBundle::initializePlatformDefaults):
1360         Add stubs for initializePlatformDefaults.
1361
1362         * WebKitTestRunner/TestController.cpp:
1363         (WTR::TestController::runTest):
1364         (WTR::TestController::runTestingServerLoop):
1365         (WTR::TestController::run):
1366         Parse expected pixel results out of stdin and setup the invocation with it.
1367
1368         * WebKitTestRunner/TestInvocation.cpp:
1369         (WTR::TestInvocation::TestInvocation):
1370         Initialize new members.
1371
1372         (WTR::TestInvocation::~TestInvocation):
1373         (WTR::TestInvocation::setIsPixelTest):
1374         (WTR::sizeWebViewForCurrentTest):
1375         (WTR::TestInvocation::invoke):
1376         (WTR::TestInvocation::dump):
1377         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
1378         * WebKitTestRunner/TestInvocation.h:
1379         Dump pixels in addition to text by calling dumpPixelsAndCompareWithExpected.
1380
1381         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
1382         (WTR::PlatformWebView::PlatformWebView):
1383         Use -[NSColorSpace genericRGBColorSpace] instead of the main screen color space.
1384
1385         * WebKitTestRunner/mac/TestInvocationMac.mm: Added.
1386         (WTR::createCGContextFromPlatformView):
1387         (WTR::computeMD5HashStringForContext):
1388         (WTR::dumpBitmap):
1389         (WTR::forceRepaintFunction):
1390         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
1391         Add mac specific pixel dumping support. Right now, we always pull pixels
1392         from the window server.
1393
1394         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
1395         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
1396         * WebKitTestRunner/qt/WebKitTestRunner.pro:
1397         * WebKitTestRunner/win/InjectedBundle.vcproj:
1398         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
1399         Add new files.
1400
1401 2011-02-24  Chang Shu  <cshu@webkit.org>
1402
1403         Reviewed by Kenneth Rohde Christiansen.
1404
1405         [Qt] Launch MiniBrowser if the test run is WebKit2. 
1406         https://bugs.webkit.org/show_bug.cgi?id=55145
1407
1408         * Scripts/old-run-webkit-tests:
1409         * Scripts/run-launcher:
1410         * Scripts/webkitdirs.pm:
1411
1412 2011-02-24  Victoria Kirst  <vrk@google.com>
1413
1414         Reviewed by Mihai Parparita.
1415
1416         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
1417         https://bugs.webkit.org/show_bug.cgi?id=55089
1418
1419         This adds an alias into the media/ directory so that http/tests/media
1420         tests can access the media resources when running in an httpd process.
1421
1422         * Scripts/webkitperl/httpd.pm:
1423         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1424         * Scripts/webkitpy/layout_tests/port/http_server.py:
1425
1426 2011-02-24  Andrew Wilson  <atwilson@chromium.org>
1427
1428         Unreviewed, rolling out r79570.
1429         http://trac.webkit.org/changeset/79570
1430         https://bugs.webkit.org/show_bug.cgi?id=54874
1431
1432         Breaks chromium build because glue/mocks/mock_web_frame.h/cc
1433         was not updated
1434
1435         * DumpRenderTree/chromium/TestShell.cpp:
1436         (TestShell::resetTestController):
1437         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1438         (resetDefaultsToConsistentValues):
1439         * DumpRenderTree/mac/DumpRenderTree.mm:
1440         (resetWebViewToConsistentStateBeforeTesting):
1441         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1442         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1443         * DumpRenderTree/win/DumpRenderTree.cpp:
1444         (resetWebViewToConsistentStateBeforeTesting):
1445         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1446         (WTR::InjectedBundlePage::reset):
1447
1448 2011-02-24  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
1449
1450         Reviewed by Laszlo Gombos.
1451
1452         [Qt] add android 2.2 user agent to QtTestBrowser
1453         https://bugs.webkit.org/show_bug.cgi?id=55085
1454
1455         Credit to Forrest Hodgkins to find right user-agent which works for youtube.
1456         * QtTestBrowser/useragentlist.txt:
1457
1458 2011-02-24  Adam Roben  <aroben@apple.com>
1459
1460         Change FrameLoadDelegate to support any number of delegates with delayed work to process
1461
1462         This makes our behavior match Mac more closely, and allows us to remove an incorrect
1463         assertion that was firing during some tests. (The assertion was claiming that there was
1464         never more than one delegate with delayed work to process, but that was not the case.)
1465
1466         Fixes <http://webkit.org/b/55146> Assertion failure in FrameLoadDelegate::locationChangeDone
1467         when running http/tests/navigation/back-twice-without-commit.html
1468
1469         Reviewed by Eric Carlson.
1470
1471         * DumpRenderTree/win/FrameLoadDelegate.cpp:
1472         (delegatesWithDelayedWork): Added. Returns all FrameLoadDelegates that have delayed work to
1473         process. A single delegate may appear in this Vector more than once (just as, on Mac, a
1474         single delegate may have multiple performSelector requests).
1475         (processWorkTimer): Pass the HWND to ::KillTimer, for pedantic brownie points. Added an
1476         assertion that the timer firing is the shared process work timer. Instead of using the
1477         single, global "delegate waiting for timer" delegate, give all delegates that have delayed
1478         work to process a chance to process their work.
1479         (FrameLoadDelegate::locationChangeDone): If we don't already have an active timer for
1480         processing delayed work, create one. Then add ourselves to the delegatesWithDelayedWork
1481         Vector so our processWork function will be called when the timer fires.
1482
1483 2011-02-24  Vsevolod Vlasov  <vsevik@chromium.org>
1484
1485         Reviewed by Alexey Proskuryakov.
1486
1487         DumpRenderTree should reset frame opener between tests.
1488         https://bugs.webkit.org/show_bug.cgi?id=54874
1489
1490         Added clearOpener() calls between tests.
1491         Some tests are using window.open() to change current window location.
1492         This changes window.opener property whose nullity is checked before
1493         showing xml viewer.
1494
1495         * DumpRenderTree/chromium/TestShell.cpp:
1496         (TestShell::resetTestController):
1497         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1498         (resetDefaultsToConsistentValues):
1499         * DumpRenderTree/mac/DumpRenderTree.mm:
1500         (resetWebViewToConsistentStateBeforeTesting):
1501         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1502         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1503         * DumpRenderTree/win/DumpRenderTree.cpp:
1504         (resetWebViewToConsistentStateBeforeTesting):
1505         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1506         (WTR::InjectedBundlePage::reset):
1507
1508 2011-02-24  Adam Roben  <aroben@apple.com>
1509
1510         Don't trigger Windows builds for changes to WebCore.exp.in or the top-level GNUmakefile.am
1511
1512         Fixes <http://webkit.org/b/55144> Windows builders should not have built r79343 or r79440,
1513         but did
1514
1515         Reviewed by Anders Carlsson.
1516
1517         * Scripts/webkitpy/common/config/build.py:
1518         (_should_file_trigger_build): Loosened the pattern for GNUmakefile.am to include the
1519         top-level GNUmakefile.am, and loosened the pattern for .exp files to include WebCore.exp.in.
1520
1521         * Scripts/webkitpy/common/config/build_unittest.py:
1522         (ShouldBuildTest): Added some more tests.
1523
1524 2011-02-24  Andras Becsi  <abecsi@webkit.org>
1525
1526         Reviewed by Laszlo Gombos.
1527
1528         [Qt] MinGW build fails to link
1529         https://bugs.webkit.org/show_bug.cgi?id=55050
1530
1531         Prepend the libraries of subcomponents instead of appending them
1532         to fix the library order according to the dependency of the libraries
1533
1534         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
1535         * WebKitTestRunner/qt/WebKitTestRunner.pro:
1536
1537 2011-02-24  Gabor Loki  <loki@webkit.org>
1538
1539         Reviewed by Csaba Osztrogonác.
1540
1541         [Qt] Remove ARMv5 Release buildbot.
1542
1543         Remove ARMv5 Release buildbot, because there is very little difference
1544         between ARMv5 and ARMv7 Release bots, and there is also very little
1545         interest in ARMv5 build slave (compared to ARMv7).
1546
1547         * BuildSlaveSupport/build.webkit.org-config/config.json:
1548         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1549
1550 2011-02-22  Ojan Vafai  <ojan@chromium.org>
1551
1552         Reviewed by Tony Chang.
1553
1554         fix revision number finding when creating git patches
1555         https://bugs.webkit.org/show_bug.cgi?id=55024
1556
1557         HEAD~n does not walk merge commits correctly. git log however does.
1558
1559         * Scripts/webkitpy/common/checkout/scm.py:
1560         * Scripts/webkitpy/common/checkout/scm_unittest.py:
1561
1562 2011-02-23  Kenneth Russell  <kbr@google.com>
1563
1564         Unreviewed, rolling out r79387.
1565         https://bugs.webkit.org/show_bug.cgi?id=54885
1566
1567         Reapply the code changes from r79038 since they weren't the root
1568         cause of the test timeouts.
1569
1570         * DumpRenderTree/chromium/WebViewHost.cpp:
1571         (WebViewHost::reset):
1572
1573 2011-02-23  Siddharth Mathur  <siddharth.mathur@nokia.com>
1574
1575         Reviewed by Laszlo Gombos.
1576
1577         [Qt] Make sure Symbian binary UIDs are unique
1578
1579         * MiniBrowser/qt/MiniBrowser.pro: Resolve the UID collision with
1580         QtTestBrowser.pro by changing the UID.
1581
1582 2011-02-23  Patrick Gansterer  <paroga@webkit.org>
1583
1584         Reviewed by Darin Adler.
1585
1586         Rename PLATFORM(CF) to USE(CF)
1587         https://bugs.webkit.org/show_bug.cgi?id=53540
1588
1589         * DumpRenderTree/config.h:
1590
1591 2011-02-23  Dirk Pranke  <dpranke@chromium.org>
1592
1593         Reviewed by Ojan Vafai.
1594
1595         nrwt: turn on multiprocessing on chromium_linux, clean up unit tests for worker model defaults
1596         https://bugs.webkit.org/show_bug.cgi?id=55016
1597
1598         * Scripts/webkitpy/layout_tests/port/base.py:
1599         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1600         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py: Added.
1601         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1602         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1603         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1604         * Scripts/webkitpy/layout_tests/port/mac.py:
1605         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1606         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
1607         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1608
1609 2011-02-23  Tony Chang  <tony@chromium.org>
1610
1611         Reviewed by Kent Tamura.
1612
1613         [chromium] DumpRenderTree --check-layout-test-sys-deps leaks 4 temp dirs
1614         https://bugs.webkit.org/show_bug.cgi?id=55004
1615
1616         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1617         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
1618         (WebKitSupportTestEnvironment::~WebKitSupportTestEnvironment):
1619         (main): Use a C++ object to ensure that TearDownTestEnvironment
1620             gets called when --check-layout-test-sys-deps is called
1621
1622 2011-02-23  Gopal Raghavan  <gopal.1.raghavan@nokia.com>
1623
1624         Reviewed by Anders Carlsson.
1625
1626         [Qt] [Qt] check webkit style fails in qt/tests
1627         https://bugs.webkit.org/show_bug.cgi?id=54998
1628
1629         check-webkit-style was complaining about missing config.h in WebKit/qt/test/qdeclarativewebview. This patch was originally part of 50222. Reviewer suggested to split and create a separate patch for checker.
1630
1631         * Scripts/webkitpy/style/checker.py:
1632
1633 2011-02-22  Vsevolod Vlasov  <vsevik@chromium.org>
1634
1635         Reviewed by Pavel Feldman.
1636
1637         DumpRenderTree should dump as text if test path contains "dumpAsText/"
1638         https://bugs.webkit.org/show_bug.cgi?id=54873
1639
1640         * DumpRenderTree/chromium/LayoutTestController.h:
1641         (LayoutTestController::setShouldDumpAsText):
1642         (LayoutTestController::setShouldGeneratePixelResults):
1643         * DumpRenderTree/chromium/TestShell.cpp:
1644         (TestShell::runFileTest):
1645         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1646         (shouldDumpAsText):
1647         (dump):
1648         (runTest):
1649         * DumpRenderTree/mac/DumpRenderTree.mm:
1650         (shouldDumpAsText):
1651         (runTest):
1652         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1653         (WebCore::isDumpAsTextTest):
1654         (WebCore::DumpRenderTree::open):
1655         * DumpRenderTree/win/DumpRenderTree.cpp:
1656         (shouldDumpAsText):
1657         (runTest):
1658         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1659         (WTR::InjectedBundlePage::dump):
1660
1661 2011-02-22  Kenneth Russell  <kbr@google.com>
1662
1663         Reviewed by James Robinson.
1664
1665         Many tests started timing out on the chrome mac canaries around r79035
1666         https://bugs.webkit.org/show_bug.cgi?id=54885
1667
1668         Revert the code changes, but not the Chromium DEPS roll, from
1669         r79038 to see whether this clears up the timing-out tests. Was not
1670         able to reproduce the timeouts locally with a Release mode build
1671         of DRT (on Snow Leopard).
1672
1673         * DumpRenderTree/chromium/WebViewHost.cpp:
1674         (WebViewHost::reset):
1675
1676 2011-02-22  Dirk Pranke  <dpranke@chromium.org>
1677
1678         Reviewed by James Robinson.
1679
1680         nrwt: --platform chromium-gpu doesn't work with --worker-model=processes
1681         https://bugs.webkit.org/show_bug.cgi?id=55009
1682
1683         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1684
1685 2011-02-18  Mihai Parparita  <mihaip@chromium.org>
1686
1687         Reviewed by Ojan Vafai.
1688
1689         Default to 500 for --exit-after-n-failures
1690         https://bugs.webkit.org/show_bug.cgi?id=54773
1691
1692         Make 500 the default value for --exit-after-n-failures (ORWT bots
1693         already default to this as of r75726). Similarly, change the default for
1694         --exit-after-n-crashes-or-timeouts to 20.
1695
1696         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1697
1698 2011-02-02  Martin Robinson  <mrobinson@igalia.com>
1699
1700         Reviewed by Xan Lopez.
1701
1702         [GTK] DRT needs an implementation of LayoutTestController.setSelectTrailingWhitespace
1703         https://bugs.webkit.org/show_bug.cgi?id=53603
1704
1705         Add support for LayoutTestController.setSelectTrailingWhitespace by calling through
1706         to DumpRenderTreeSupportGtk for this functionality.
1707
1708         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1709         (resetDefaultsToConsistentValues): Always call LayoutTestController.setSelectTrailingWhitespace to
1710         false when reseting values between tests.
1711         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1712         (LayoutTestController::setSelectTrailingWhitespaceEnabled): Call LayoutTestController.setSelectTrailingWhitespace
1713         with the appropriate value.
1714
1715 2011-02-22  Adam Roben  <aroben@apple.com>
1716
1717         Don't trigger a Windows build when the top-level ChangeLog changes
1718
1719         r79320 is an example of a change that shouldn't have built on Windows, but did, because of
1720         this bug.
1721
1722         Fixes <http://webkit.org/b/54957> Windows bots build when top-level ChangeLog changes, but
1723         shouldn't
1724
1725         Reviewed by Anders Carlsson.
1726
1727         * Scripts/webkitpy/common/config/build.py:
1728         (_should_file_trigger_build): Modified the ChangeLog pattern to match the top-level
1729         ChangeLog, and re-sorted it in the list.
1730
1731         * Scripts/webkitpy/common/config/build_unittest.py:
1732         (ShouldBuildTest): Added a test that ChangeLogs don't trigger builds.
1733
1734 2011-02-22  Sheriff Bot  <webkit.review.bot@gmail.com>
1735
1736         Unreviewed, rolling out r78691.
1737         http://trac.webkit.org/changeset/78691
1738         https://bugs.webkit.org/show_bug.cgi?id=54956
1739
1740         Leopard Debug is successfully building again (Requested by
1741         aroben on #webkit).
1742
1743         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1744         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1745
1746 2011-02-22  Ilya Tikhonovsky  <loislo@chromium.org>
1747
1748         Reviewed by Pavel Feldman.
1749
1750         Web Inspector: flakyness of inspector tests.
1751         https://bugs.webkit.org/show_bug.cgi?id=54729
1752
1753         As far as we have the protocol with sequence numbers we can simplify test harness support
1754         and drop out chomium specific methods.
1755
1756         * DumpRenderTree/DumpRenderTree.gypi:
1757         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
1758         (DRTDevToolsAgent::sendMessageToInspectorFrontend):
1759         (DRTDevToolsAgent::runtimePropertyChanged):
1760         (DRTDevToolsAgent::asyncCall):
1761         (DRTDevToolsAgent::call):
1762         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
1763         * DumpRenderTree/chromium/DRTDevToolsCallArgs.cpp: Removed.
1764         * DumpRenderTree/chromium/DRTDevToolsCallArgs.h: Removed.
1765         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
1766         (DRTDevToolsClient::sendMessageToBackend):
1767         (DRTDevToolsClient::asyncCall):
1768         (DRTDevToolsClient::call):
1769         * DumpRenderTree/chromium/DRTDevToolsClient.h:
1770
1771 2011-02-22  Andras Becsi  <abecsi@webkit.org>
1772
1773         Reviewed by Laszlo Gombos.
1774
1775         [Qt] Redesign the build system
1776         https://bugs.webkit.org/show_bug.cgi?id=51339
1777
1778         Part 2.
1779
1780         Build WebCore as a static library, compile the WebKit API and WebKit2 API
1781         in a final step and link to WebKit2, WebCore and JSC libraries to fix
1782         linking issues resulting from stripped away symbols.
1783
1784         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Remove superfluous includepaths.
1785
1786 2011-02-21  Lukasz Slachciak  <l.slachciak@samsung.com>
1787
1788         Reviewed by Antonio Gomes.
1789
1790         [EFL] Restored spatial navigation option in EWebLauncher since #18662 bug is closed
1791         https://bugs.webkit.org/show_bug.cgi?id=54806
1792
1793         * EWebLauncher/main.c:
1794         (on_key_down): calling ewk_view_setting_spatial_navigation_set when "F12" pressed added.
1795
1796 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
1797
1798         Reviewed by Kent Tamura.
1799
1800         nrwt: revert behavior on 10.5 to "old-threads". It's possible
1801         that r79062 has introduced a bunch of flakiness, although I
1802         can't explain why right now. This change should restore the
1803         old behavior, and then we can look at the build logs to see
1804         if the tests stop timing out or being flaky.
1805
1806         https://bugs.webkit.org/show_bug.cgi?id=54925
1807
1808         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1809         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1810         * Scripts/webkitpy/layout_tests/port/mac.py:
1811         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1812
1813 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
1814
1815         Reviewed by Dimitri Glazkov.
1816
1817         nrwt: fix typo causing us to try and repeatedly shut down the servers
1818         https://bugs.webkit.org/show_bug.cgi?id=54904
1819
1820         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
1821
1822 2011-02-21  Alejandro G. Castro  <alex@igalia.com>
1823
1824         Reviewed by Csaba Osztrogonác.
1825
1826         REGRESSION(r79157):
1827         plugins/get-url-with-javascript-destroying-plugin.html make the
1828         next test crash
1829         https://bugs.webkit.org/show_bug.cgi?id=54863
1830
1831         * GNUmakefile.am:
1832
1833 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
1834
1835         Reviewed by Ojan Vafai.
1836
1837         test-webkitpy: stop skipping a bunch of tests on win32
1838         https://bugs.webkit.org/show_bug.cgi?id=54788
1839
1840         * Scripts/webkitpy/common/net/testoutputset_unittest.py:
1841         * Scripts/webkitpy/test/main.py:
1842
1843 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
1844
1845         Reviewed by Ojan Vafai.
1846
1847         Fix various bugs related to handling of dir separators and real
1848         vs. fake filesystems that were causing test-webkitpy to fail
1849         tests on win32. These bugs have a variety of causes but are
1850         lumped together because they should be easy to review.
1851
1852         https://bugs.webkit.org/show_bug.cgi?id=54700
1853
1854         * Scripts/webkitpy/common/net/testoutput.py:
1855           Here we were using os.path.sep but not escaping it prior to
1856           using in a regexp, which  caused bad things on win32.
1857         * Scripts/webkitpy/common/net/testoutput_unittest.py:
1858           Same here.
1859         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1860           Here we were using the native filesystem instead of the mock
1861           filesystem, making some tests unnecessarily platform dependent.
1862           Also we weren't setting the default configuration for the
1863           TestChromiumWinPort.
1864         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py:
1865           Here we *shouldn't* use os.path.join(), just so the code becomes
1866           less platform dependent.
1867         * Scripts/webkitpy/style/checker.py:
1868           Here we should be using os.path.sep instead of "/"; the tests
1869           require the native filesystem.
1870         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1871           Here the underlying code is using os.path.sep, so we probably
1872           should as well.
1873         * Scripts/webkitpy/style/test_expectations_unittest.py:
1874           Here we should be using the filesystem hanging of the port
1875           object (which is the mocked filesystem) rather than the native
1876           one.
1877
1878 2011-02-20  Anders Carlsson  <andersca@apple.com>
1879
1880         Reviewed by Maciej Stachowiak.
1881
1882         Crash when a plug-in requests a javascript: url that destroys the plug-in
1883         https://bugs.webkit.org/show_bug.cgi?id=54837
1884         <rdar://problem/9005475>
1885
1886         Add new plug-in test.
1887
1888         * DumpRenderTree/DumpRenderTree.gypi:
1889         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1890         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
1891         (PluginTest::NPN_GetURL):
1892         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
1893         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURLDestroyingPlugin.cpp: Added.
1894         (GetURLWithJavaScriptURLDestroyingPlugin::GetURLWithJavaScriptURLDestroyingPlugin):
1895         (GetURLWithJavaScriptURLDestroyingPlugin::NPP_New):
1896         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
1897         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1898
1899 2011-02-19  Patrick Gansterer  <paroga@webkit.org>
1900
1901         Reviewed by Andreas Kling.
1902
1903         Make WinCE a core builder
1904         https://bugs.webkit.org/show_bug.cgi?id=54801
1905
1906         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1907         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1908
1909 2011-02-19  Lukasz Slachciak  <l.slachciak@samsung.com>
1910
1911         Reviewed by Andreas Kling.
1912
1913         Fixed problem with launching EWebLauncher with "run-launcher --efl". Incorrect path referring to non-existing .libs directory ws set.
1914         https://webkit.org/b/54778
1915
1916         * Scripts/webkitdirs.pm:
1917
1918 2011-02-18  David Levin  <levin@chromium.org>
1919
1920         Reviewed by Eric Seidel.
1921
1922         check-webkit-style: Misses brace style error when the line with the { has a } in it.
1923         https://bugs.webkit.org/show_bug.cgi?id=54769
1924
1925         * Scripts/webkitpy/style/checkers/cpp.py: Fixed the check for the close brace
1926           to only look after the last open brace, so that the open brace in this line
1927           "} else {" will still be able to trigger the error.
1928         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added related unit tests.
1929
1930 2011-02-18  Zan Dobersek  <zandobersek@gmail.com>
1931
1932         Reviewed by Martin Robinson.
1933
1934         [Gtk] Re-enable meter tag support
1935         https://bugs.webkit.org/show_bug.cgi?id=54762
1936
1937         Meter tag is supported on the Gtk port, enable it by default.
1938
1939         * Scripts/build-webkit:
1940
1941 2011-02-18  David Levin  <levin@chromium.org>
1942
1943         Reviewed by Eric Seidel.
1944
1945         check-webkit-style falsely complains about WebKitGTK+ public headers
1946         https://bugs.webkit.org/show_bug.cgi?id=54650
1947
1948         * Scripts/webkitpy/style/checker.py: Add some exceptions for the gtk files.
1949         * Scripts/webkitpy/style/checkers/cpp.py: Restrict the checks for WEBKIT_API
1950           to the chromium directory (and improved the checks slightly).
1951         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added corresponding tests.
1952
1953 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
1954
1955         Reviewed by James Robinson.
1956
1957         new-run-webkit-tests: enable multiple processes by default on
1958         mac. This change removes the artificial restrictions we placed
1959         on the # of child processes we used with the old threading
1960         model, and switches to the new message-based model and multiple
1961         processes, where available. If multiple processes are not
1962         available (Leopard / Python 2.5), then we use the 'inline' model
1963         instead of the 'old-threads' model on the 'mac' port or one
1964         process on the 'chromium-mac' port. We need additional testing
1965         to see if the new 'threads' model works reliably and is worth
1966         supporting, or if we should just live with things being slightly
1967         slow.
1968
1969         https://bugs.webkit.org/show_bug.cgi?id=54596
1970
1971         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1972         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1973         * Scripts/webkitpy/layout_tests/port/mac.py:
1974         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1975         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
1976
1977 2011-02-18  Andrew Wilson  <atwilson@chromium.org>
1978
1979         Unreviewed, rolling out r79047.
1980         http://trac.webkit.org/changeset/79047
1981         https://bugs.webkit.org/show_bug.cgi?id=54596
1982
1983         Broke canary bots - please watch the canaries the next time
1984         you land this
1985
1986         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1987         * Scripts/webkitpy/layout_tests/port/mac.py:
1988
1989 2011-02-18  Emil A Eklund  <eae@chromium.org>
1990
1991         Reviewed by Darin Adler.
1992
1993         Crash in EventHandler::sendContextMenuEventForKey
1994         https://bugs.webkit.org/show_bug.cgi?id=54495
1995
1996         Add support for the context menu key (VK_APPS) to EventSender for
1997         platforms that support the key.
1998
1999         * DumpRenderTree/chromium/EventSender.cpp:
2000         (EventSender::keyDown):
2001         * DumpRenderTree/gtk/EventSender.cpp:
2002         (keyDownCallback):
2003         * DumpRenderTree/qt/EventSenderQt.cpp:
2004         (EventSender::keyDown):
2005         * DumpRenderTree/win/EventSender.cpp:
2006         (keyDownCallback):
2007
2008 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2009
2010         Reviewed by Ojan Vafai.
2011
2012         new-run-webkit-tests: enable multiple processes by default on mac
2013         https://bugs.webkit.org/show_bug.cgi?id=54596
2014
2015         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2016         * Scripts/webkitpy/layout_tests/port/mac.py:
2017
2018 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2019
2020         Reviewed by Adam Roben.
2021
2022         Skip pretty patch unit tests if ruby isn't installed.
2023
2024         https://bugs.webkit.org/show_bug.cgi?id=54699
2025
2026         * Scripts/webkitpy/common/prettypatch_unittest.py:
2027
2028 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2029
2030         Reviewed by Adam Roben.
2031
2032         Skip mac.test_skipped_file_paths() when running on win32; the
2033         test has hardcoded '/' directory separators and since it's a
2034         port-specific test and we have coverage on other ports it's not
2035         worth it to make the test more generic.
2036
2037         https://bugs.webkit.org/show_bug.cgi?id=54698
2038
2039         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2040
2041 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2042
2043         Reviewed by Adam Roben.
2044
2045         Fix webkitpy.layout_tests.port.config_unittest to work on Win32.
2046
2047         https://bugs.webkit.org/show_bug.cgi?id=54703
2048
2049         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
2050
2051 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2052
2053         Reviewed by Adam Roben.
2054
2055         Change the default port we're mocking out on win32 when we
2056         run mock_drt_unittest. Normally we'd default to the
2057         port-specific default, but that would be "win", which doesn't
2058         currently work.
2059
2060         https://bugs.webkit.org/show_bug.cgi?id=54708
2061
2062         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2063
2064 2011-02-18  Kenneth Russell  <kbr@google.com>
2065
2066         Reviewed by James Robinson.
2067
2068         [chromium] Clean up initialization of minimum timer interval
2069         https://bugs.webkit.org/show_bug.cgi?id=54772
2070
2071         Use new GetForegroundTabTimerInterval function in webkit_support
2072         to reset the minimum timer interval before each test.
2073
2074         * DumpRenderTree/chromium/WebViewHost.cpp:
2075         (WebViewHost::reset):
2076
2077 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2078
2079         Reviewed by Adam Roben.
2080
2081         r78494 fixed the chromium_win unittest failing under win32,
2082         but not cygwin. This fixes cygwin as well.
2083
2084         https://bugs.webkit.org/show_bug.cgi?id=54652
2085
2086         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2087
2088 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2089
2090         Reviewed by Ojan Vafai, Mihai Parparita.
2091
2092         new-run-webkit-tests: r78522 made it impossible to create a
2093         'chromium-win' or 'chromium-mac' port; they would always get the
2094         version tacked on to the end. It turns out that we actually need
2095         to be able to create these ports in order for
2096         rebaseline-chromium-webkit-tests to work correctly given the
2097         broken coverage of our layout bots. As soon as we get SL and
2098         Win 7 bots running reliably on the canaries, we can get rid of
2099         this change. Note that we only need to change the chromium_mac
2100         and win ports because linux doesn't yet use versions and that
2101         script doesn't work with non-chromium ports.
2102
2103         Also, this change fixes a bug where we were using
2104         "platform/chromium-win-win7" and
2105         "platform/chromium-mac-snowleopard" as the baseline dirs for
2106         Win 7 and SL; we should've been using the generic directories
2107         instead since they are the newest versions of the os's.
2108
2109         https://bugs.webkit.org/show_bug.cgi?id=54691
2110
2111         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2112         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2113         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2114         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2115         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2116
2117 2011-02-16  MORITA Hajime  <morrita@google.com>
2118
2119         Reviewed by Kent Tamura.
2120
2121         [Chromium][DRT] some PostDelayedTask() accesses possibly dangling this pointer.
2122         http://webkit.org/b/53899
2123
2124         Replaced static invocation wrappers with MethodTask objects,
2125         which are removed on WebViewHost destructor.
2126         
2127         * DumpRenderTree/chromium/WebViewHost.cpp:
2128         (WebViewHost::requestCheckingOfText):
2129         (WebViewHost::scheduleAnimation):
2130         (WebViewHost::closeWidgetSoon):
2131         * DumpRenderTree/chromium/WebViewHost.h:
2132         (WebViewHost::HostMethodTask::HostMethodTask):
2133         (WebViewHost::HostMethodTask::runIfValid):
2134         (WebViewHost::taskList):
2135
2136 2011-02-07  Hayato Ito  <hayato@chromium.org>
2137
2138         Reviewed by Ojan Vafai.
2139
2140         Get rid of code which writes test results from test_type's
2141         compare_output() method.
2142         Writing test results now happens in single_test_runner, using
2143         newly introduced test_result_writer module.
2144
2145         Subsequent patches will eliminate test_type/* classes.
2146
2147         https://bugs.webkit.org/show_bug.cgi?id=53908
2148
2149         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2150         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py: Added.
2151         * Scripts/webkitpy/layout_tests/test_types/image_diff.py:
2152         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
2153         * Scripts/webkitpy/layout_tests/test_types/text_diff.py:
2154
2155 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2156
2157         Reviewed by Adam Roben.
2158
2159         Add blacklist of remaining webkitpy/* modules that fail on
2160         Win32, so that test-webkitpy will at least run cleanly.
2161
2162         https://bugs.webkit.org/show_bug.cgi?id=54709
2163
2164         * Scripts/webkitpy/layout_tests/test/main.py:
2165
2166 2011-02-17  Simon Fraser  <simon.fraser@apple.com>
2167
2168         Reviewed by Sam Weinig.
2169
2170         Set a preference to ensure that scrollbars are in a predictable
2171         state for DRT.
2172
2173         * DumpRenderTree/mac/DumpRenderTree.mm:
2174         (resetDefaultsToConsistentValues):
2175
2176 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2177
2178         Reviewed by Adam Roben.
2179
2180         test-webkitpy doesn't work with the multiprocessing module on win32
2181         or cygwin, so we skip it for now.
2182
2183         https://bugs.webkit.org/show_bug.cgi?id=54520
2184
2185         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
2186         * Scripts/webkitpy/layout_tests/layout_package/run_webkit_tests_unittest.py:
2187
2188 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2189
2190         Reviewed by Adam Roben.
2191
2192         Fix version, baseline_search_path for chromium-win-win7. Also
2193         fix chromium_win_unittests, and fix port_testcase.make_port()
2194         to correctly use the options passed into it and only set
2195         results_directory() if necessary.
2196
2197         https://bugs.webkit.org/show_bug.cgi?id=54652
2198
2199         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2200         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2201         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2202
2203 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2204
2205         Reviewed by Ojan Vafai.
2206
2207         Fix a hang in new-run-webkit-tests when run with
2208         --dry-run --worker-model=processes
2209
2210         https://bugs.webkit.org/show_bug.cgi?id=54595
2211
2212         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
2213         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2214
2215 2011-02-17  Kevin Ollivier  <kevino@theolliviers.com>
2216
2217         [wx] Build fixes after recent changes.
2218
2219         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2220         (LayoutTestController::setMinimumTimerInterval):
2221         * wx/build/settings.py:
2222
2223 2011-02-17  Csaba Osztrogonác  <ossy@webkit.org>
2224
2225         Reviewed by Andreas Kling.
2226
2227         [Qt] ImageDiff sometimes hangs
2228         https://bugs.webkit.org/show_bug.cgi?id=54641
2229
2230         * DumpRenderTree/qt/ImageDiff.cpp: Clear actualImage and baselineImage after all comparison.
2231         (main):
2232
2233 2011-02-17  Peter Varga  <pvarga@webkit.org>
2234
2235         Reviewed by Csaba Osztrogonác.
2236
2237         The run-sunspider --v8-suite doesn't work
2238         https://bugs.webkit.org/show_bug.cgi?id=54660
2239
2240         * Scripts/run-sunspider:
2241
2242 2011-02-17  Gabor Rapcsanyi  <rgabor@webkit.org>
2243
2244         Reviewed by Csaba Osztrogonác.
2245
2246         [NRWT] Fix environment setup on Qt port.
2247         https://bugs.webkit.org/show_bug.cgi?id=54549
2248
2249         * Scripts/webkitpy/layout_tests/port/qt.py:
2250
2251 2011-02-16  Yael Aharon  <yael.aharon@nokia.com>
2252
2253         Reviewed by Andreas Kling.
2254
2255         [Qt] Enable tiled backing store by default in MiniBrowser.
2256         https://bugs.webkit.org/show_bug.cgi?id=54581
2257
2258         Instead of a flag to turn it on, a new flag is defined to turn
2259         tiled backing store off.
2260
2261         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2262         (MiniBrowserApplication::handleUserOptions):
2263         * MiniBrowser/qt/MiniBrowserApplication.h:
2264         (WindowOptions::WindowOptions):
2265
2266 2011-02-16  Eric Seidel  <eric@webkit.org>
2267
2268         Reviewed by Mihai Parparita.
2269
2270         flaky test archive uploads are always 22 bytes long
2271         https://bugs.webkit.org/show_bug.cgi?id=54593
2272
2273         Speculative fix for the 22-bytes bug.
2274
2275         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
2276
2277 2011-02-16  Anna Cavender  <annacc@chromium.org>
2278
2279         Reviewed by Mihai Parparita.
2280
2281         http/tests/media should be served over HTTP (not from local file)
2282         https://bugs.webkit.org/show_bug.cgi?id=54028
2283
2284         Removed special file:// treatment of http/tests/media
2285
2286         * Scripts/old-run-webkit-tests:
2287         * Scripts/webkitpy/layout_tests/port/base.py:
2288
2289 2011-02-16  Mihai Parparita  <mihaip@chromium.org>
2290
2291         Reviewed by Ojan Vafai.
2292
2293         ChangeLog.parse_latest_entry_from_file does not handle rolled over ChangeLogs
2294         https://bugs.webkit.org/show_bug.cgi?id=54609
2295         
2296         r78737 rolled over ChangeLogs, and webkit-patch complains when landing
2297         patches such as this one that are the first to add an entry to the
2298         ChangeLog (since they don't see a second date line to know where the
2299         entry ends).
2300
2301         * Scripts/webkitpy/common/checkout/changelog.py:
2302         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
2303
2304 == Rolled over to ChangeLog-2011-02-16 ==