a5c076261f00985672c5dd4f57e130c9babe7a33
[WebKit-https.git] / Tools / ChangeLog
1 2011-03-03  Mihai Parparita  <mihaip@chromium.org>
2
3         Reviewed by Tony Chang.
4
5         NRWT: AttributeError: TestRunner2 instance has no attribute '_cancel_workers'
6         https://bugs.webkit.org/show_bug.cgi?id=55694
7         
8         Fixes for interruptions in NRWT:
9         - Make TestRunInterruptedException be pickleable correctly (the base
10           Exception class defines a __reduce__ that does not include the reason)
11         - Fix ordering of returned arguments from TestRunner2._run_tests
12           (interrupted and keyboard_interrupted were reversed)
13         - Fix cancel_workers callsites (was using old name).
14         - In handle_exception re-raise actual exception instance that was thrown
15
16         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
17         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
18         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
19         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
20
21 2011-03-03  Qi Zhang  <qi.2.zhang@nokia.com>
22
23         Unreviewed.
24
25         Adding myself as a committer.
26
27         * Scripts/webkitpy/common/config/committers.py:
28
29 2011-03-03  Tony Gentilcore  <tonyg@chromium.org>
30
31         Reviewed by Mihai Parparita.
32
33         Teach sheriffbot to answer whois commands
34         https://bugs.webkit.org/show_bug.cgi?id=55687
35
36         * Scripts/webkitpy/tool/bot/irc_command.py:
37         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
38         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
39
40 2011-03-03  Victoria Kirst  <vrk@google.com>
41
42         Reviewed by Eric Carlson.
43
44         Enable media tests in chromium-gpu/test_expectations
45         https://bugs.webkit.org/show_bug.cgi?id=53608
46
47         Add media folder to the list of GPU test paths.
48
49         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
50
51 2011-03-03  Adam Roben  <aroben@apple.com>
52
53         Work around a bug in KURL's parsing of Windows-style absolute file: URLs
54
55         Fixes <http://webkit.org/b/55674> Many tests fail in WebKit2 mode on Windows due to
56         incorrect parsing of absolute Windows-style file: URLs
57
58         Reviewed by Tim Hatcher.
59
60         * WebKitTestRunner/TestInvocation.cpp:
61         (WTR::createWKURL): Add a "localhost" host to file: URLs we create to work around
62         <http://webkit.org/b/55683>.
63
64 2011-03-03  Tony Chang  <tony@chromium.org>
65
66         Reviewed by Dimitri Glazkov.
67
68         [chromium] Run kill-old-processes on the chromium-win bot too
69         https://bugs.webkit.org/show_bug.cgi?id=55630
70
71         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
72         * BuildSlaveSupport/win/kill-old-processes: Add wdiff.exe to the kill list.
73
74 2011-03-03  Carlos Garcia Campos  <cgarcia@igalia.com>
75
76         Reviewed by Martin Robinson.
77
78         [GTK] DRT needs implementation of EventSender.scheduleAsynchronousClick
79         https://bugs.webkit.org/show_bug.cgi?id=53960
80
81         * DumpRenderTree/gtk/EventSender.cpp:
82         (sendClick):
83         (scheduleAsynchronousClickCallback):
84
85 2011-03-03  Adam Roben  <aroben@apple.com>
86
87         Look for Windows-style paths when checking whether a test is an Inspector test
88
89         WTR part of <http://webkit.org/b/55672> <rdar://problem/9080867> All inspector tests time
90         out or crash on Windows 7 Release (WebKit2 Tests)
91
92         Reviewed by Darin Adler.
93
94         * WebKitTestRunner/TestInvocation.cpp:
95         (WTR::shouldOpenWebInspector): Also look for backslashes, since that's what we'll get on
96         Windows.
97
98 2011-03-03  Adam Roben  <aroben@apple.com>
99
100         Make Windows 7 Release (Tests) a core builder
101
102         We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
103         give it a few more days of greenness first.
104
105         Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
106
107         Reviewed by Andreas Kling.
108
109         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
110         (BuildBot): Added a regexp to match Windows 7 Release (Tests).
111
112         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
113         (BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
114         Updated the core regexps to match the buildbot module. Updated the expected results to match
115         the current set of core builders.
116
117 2011-03-03  Benjamin Poulain  <benjamin.poulain@nokia.com>
118
119         Reviewed by Andreas Kling.
120
121         Adding myself as a reviewer.
122
123         * Scripts/webkitpy/common/config/committers.py:
124
125 2011-03-03  Lukasz Slachciak  <l.slachciak@samsung.com>
126
127         Reviewed by Xan Lopez.
128
129         [GTK] Extended application cache database API and added unit tests file.
130         https://bugs.webkit.org/show_bug.cgi?id=55335
131
132         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: removed external declaration
133         of webkit_application_cache_set_maximum_size due to API exposal.
134
135 2011-03-02  Ojan Vafai  <ojan@chromium.org>
136
137         Reviewed by Adam Barth.
138
139         make webkit-patch upload work for security bugs
140         https://bugs.webkit.org/show_bug.cgi?id=55655
141
142         The problem is that we shell out to prepare-ChangeLog,
143         which does not authenticate with bugzilla. All it needs from
144         bugzilla is the bug title though. We just pass it through from
145         webkit-patch instead.
146
147         * Scripts/prepare-ChangeLog:
148         * Scripts/webkitpy/common/net/bugzilla/bug.py:
149         * Scripts/webkitpy/tool/steps/preparechangelog.py:
150
151 2011-03-01  Ojan Vafai  <ojan@chromium.org>
152
153         Reviewed by Tony Chang.
154
155         [test-dashboard] make gtest JSON match layout-tests JSON
156         https://bugs.webkit.org/show_bug.cgi?id=55556
157
158         http://trac.webkit.org/changeset/80090 stopped converting time_ms to t.
159         Make this code match.
160
161         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
162
163 2011-03-02  Dirk Pranke  <dpranke@chromium.org>
164
165         Reviewed by Tony Chang.
166
167         NRWT - implement Linux Hardy 64-bit port support. This involves
168         adding support for non-'x86' architectures and updating the
169         "deduplicate_tests" script to correctly understand that
170         port.name() is not always equal to basename(port.baseline_path()).
171
172         This change introduces two new port names: 'chromium-linux-x86'
173         and 'chromium-linux-x86_64'. Using the prior 'chromium-linux'
174         flag will cause the code to determine which one to use at
175         runtime, just like we currently do with 'chromium-win'.
176
177         https://bugs.webkit.org/show_bug.cgi?id=55535
178
179         * Scripts/webkitpy/common/system/executive_mock.py:
180           Handle 'return_stderr' keyword arg to run_command()
181         * Scripts/webkitpy/layout_tests/deduplicate_tests.py:
182         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
183         * Scripts/webkitpy/layout_tests/port/base.py:
184         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
185         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
186         * Scripts/webkitpy/layout_tests/port/factory.py:
187         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
188
189 2011-03-02  Kevin Ollivier  <kevino@theolliviers.com>
190
191         [wx] Build fix, make sure we put the wxWebKit dylib in the right directory and
192         update symlinks accordingly.
193
194         * wx/packaging/build-mac-installer.py:
195
196 2011-03-02  Sheriff Bot  <webkit.review.bot@gmail.com>
197
198         Unreviewed, rolling out r80139.
199         http://trac.webkit.org/changeset/80139
200         https://bugs.webkit.org/show_bug.cgi?id=55620
201
202         failing and crashing tests on Snow Leopard bot (Requested by
203         mihaip1 on #webkit).
204
205         * DumpRenderTree/LayoutTestController.cpp:
206         (LayoutTestController::staticFunctions):
207         * DumpRenderTree/LayoutTestController.h:
208         * DumpRenderTree/chromium/LayoutTestController.cpp:
209         (LayoutTestController::LayoutTestController):
210         * DumpRenderTree/chromium/LayoutTestController.h:
211         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
212         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
213         * DumpRenderTree/mac/UIDelegate.h:
214         * DumpRenderTree/mac/UIDelegate.mm:
215         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
216         * DumpRenderTree/qt/LayoutTestControllerQt.h:
217         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
218         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
219
220 2011-03-02  Adam Roben  <aroben@apple.com>
221
222         Wait for force a paint in WebKitTestRunner until we're actually ready to dump the output
223
224         This matches DumpRenderTree on Windows.
225
226         Fixes <http://webkit.org/b/55469> <rdar://problem/9068539> REGRESSION (r79863): Lots of
227         dom/html/level2/html/HTMLFrameElement*.html tests crashing in FrameView::paintOverhangAreas
228         in WebKit2
229
230         Reviewed by Beth Dakin.
231
232         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
233         (WTR::InjectedBundlePage::dump): Moved code to force a paint here...
234         (WTR::InjectedBundlePage::didFinishLoadForFrame): ...from here.
235
236 2011-03-02  Daniel Cheng  <dcheng@chromium.org>
237
238         Reviewed by David Levin.
239
240         Add feature define for data transfer items
241         https://bugs.webkit.org/show_bug.cgi?id=55510
242
243         * Scripts/build-webkit:
244
245 2011-03-02  Adam Roben  <aroben@apple.com>
246
247         Look in the new-to-SnowLeopard location for crash logs on SnowLeopard and newer
248
249         ReportCrash saves logs to ~/Library/Logs/CrashReporter on Leopard, but to
250         ~/Library/Logs/DiagnosticReports on SnowLeopard. old-run-webkit-tests was only looking in
251         the former location, but was getting lucky on some SnowLeopard machines because symlinks
252         were being created there pointing into DiagnosticReports. For machines without these
253         symlinks, crash logs were not getting captured.
254
255         Fixes <http://webkit.org/b/55607> old-run-webkit-tests doesn't capture crash logs on some
256         SnowLeopard machines
257
258         Reviewed by Eric Seidel.
259
260         * Scripts/old-run-webkit-tests:
261         (captureSavedCrashLog): Look in ~/Library/Logs/DiagnosticReports on SnowLeopard and newer.
262
263 2011-03-02  Adam Roben  <aroben@apple.com>
264
265         Don't try to capture crash logs on platforms that haven't implemented that feature
266
267         Fixes <http://webkit.org/b/55504> Use of uninitialized value in numeric lt (<) at
268         Tools/Scripts/old-run-webkit-tests line 1778 seen on Qt bots after a crash
269
270         Reviewed by Joseph Pecoraro.
271
272         * Scripts/old-run-webkit-tests:
273         (captureSavedCrashLog): Bail if we don't have a crash log glob.
274         (findNewestFileMatchingGlob): Fix a buggy test for when no paths match the glob. This isn't
275         strictly required to fix bug 55504, but seems worthwhile.
276
277 2011-01-26  John Knottenbelt  <jknotten@chromium.org>
278
279         Reviewed by Dmitry Titov.
280
281         Detach Geolocation from Frame when Page destroyed.
282         https://bugs.webkit.org/show_bug.cgi?id=52877
283
284         Extend the layout test controller to expose the number of pending
285         geolocation requests, so that we can test that the requests have
286         been cancelled on page close.
287
288         * DumpRenderTree/LayoutTestController.cpp:
289         (numberOfPendingGeolocationPermissionRequestsCallback):
290         (LayoutTestController::staticFunctions):
291         * DumpRenderTree/LayoutTestController.h:
292         * DumpRenderTree/chromium/LayoutTestController.cpp:
293         (LayoutTestController::LayoutTestController):
294         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
295         * DumpRenderTree/chromium/LayoutTestController.h:
296         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
297         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
298         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
299         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
300         * DumpRenderTree/mac/UIDelegate.h:
301         * DumpRenderTree/mac/UIDelegate.mm:
302         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
303         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
304         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
305         * DumpRenderTree/qt/LayoutTestControllerQt.h:
306         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
307         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
308         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
309         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
310
311
312 2011-03-01  Kent Tamura  <tkent@chromium.org>
313
314         Reviewed by Dimitri Glazkov.
315
316         Assertion fails when a form validation bubble appears
317         https://bugs.webkit.org/show_bug.cgi?id=55550
318
319         Test: fast/forms/interactive-validation-attach-assertion.html
320
321         * DumpRenderTree/chromium/WebPreferences.cpp:
322         (WebPreferences::applyTo): Enable the interactive validation feature by
323           default. This matches to Chromium browser.
324
325 2011-03-01  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
326
327         Reviewed by Kent Tamura.
328
329         Remove efl port style exception of Webcore/JavaScriptCore
330         https://bugs.webkit.org/show_bug.cgi?id=55549
331
332         WebCore and JavaScriptCore need to adhere WebKit coding style.
333
334         * Scripts/webkitpy/style/checker.py:
335
336 2011-03-01  Tony Chang  <tony@chromium.org>
337
338         Reviewed by Ojan Vafai.
339
340         [chromium] clean up temp files on chromium-mac builder too
341         https://bugs.webkit.org/show_bug.cgi?id=55525
342
343         * BuildSlaveSupport/chromium/remove-crash-logs:
344
345 2011-02-28  Ojan Vafai  <ojan@chromium.org>
346
347         Reviewed by Tony Chang.
348
349         wrap json in a function call to afford cross-domain loading
350         https://bugs.webkit.org/show_bug.cgi?id=55353
351
352         Also delete the code that compacts the JSON. It turns out this
353         breaks the rebaseline tool and is probably a premature optimization anyways.
354
355         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
356         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
357         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
358         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
359
360 2011-03-01  Sheriff Bot  <webkit.review.bot@gmail.com>
361
362         Unreviewed, rolling out r80079.
363         http://trac.webkit.org/changeset/80079
364         https://bugs.webkit.org/show_bug.cgi?id=55547
365
366         "Broke the Win debug build?" (Requested by dcheng on #webkit).
367
368         * Scripts/build-webkit:
369
370 2011-03-01  Mark Rowe  <mrowe@apple.com>
371
372         Reviewed by Sam Weinig.
373
374         Update verification scripts to handle WebKit2.
375
376         * Scripts/check-for-global-initializers: Skip object files that use RefCountedLeakChecker.
377         * Scripts/check-for-inappropriate-files-in-framework: Add some inappropriate file types for
378         WebKit2.
379
380 2011-03-01  Daniel Cheng  <dcheng@chromium.org>
381
382         Reviewed by David Levin.
383
384         Add feature define for data transfer items
385         https://bugs.webkit.org/show_bug.cgi?id=55510
386
387         * Scripts/build-webkit:
388
389 2011-02-23  Joseph Pecoraro  <joepeck@webkit.org>
390
391         Reviewed by Timothy Hatcher.
392
393         All Console Messages should be passed to ChromeClients.
394         https://bugs.webkit.org/show_bug.cgi?id=54926
395
396         Update DumpRenderTree's UIDelegate to use the new console message delegate.
397
398         * DumpRenderTree/mac/UIDelegate.mm:
399         (-[UIDelegate webView:addMessageToConsole:withSource:]):
400
401 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
402
403         Reviewed by Adam Barth.
404
405         Tragically, we stopped running the python unit tests
406         on Chromium Win just as I was getting them to be green.
407         This patch turns them back on.
408
409         https://bugs.webkit.org/show_bug.cgi?id=55521
410
411         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
412
413 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
414
415         Reviewed by Adam Roben.
416
417         Disable
418         webkitpy.common.system.executive_unittest.ExecutiveTest.\
419            test_kill_all fails on win until it can be properly triaged.
420
421         https://bugs.webkit.org/show_bug.cgi?id=54790
422
423         * Scripts/webkitpy/common/system/executive_unittest.py:
424
425     2011-03-01  Helder Correia  <helder@sencha.com>
426
427             Unreviewed. Add myself to committers.py.
428
429             * Scripts/webkitpy/common/config/committers.py:
430
431 2011-03-01  Adam Roben  <aroben@apple.com>
432
433         Give up if a crash log for the web process is taking too long to be saved on Windows
434
435         This should keep the bots from getting stuck while waiting for a crash log, if writing a
436         crash log ever hangs or the UI process for some reason doesn't notice it's finished.
437
438         Fixes <http://webkit.org/b/55499> WebKitTestRunner can hang forever waiting for a crash log
439         to be saved for the web process
440
441         Reviewed by Steve Falkenburg.
442
443         * WebKitTestRunner/win/TestControllerWin.cpp:
444         (WTR::runRunLoopUntil): Moved code to run the run loop here here from platformRunUntil.
445         Generalized the code slightly to handle the optional object to wait on.
446         (WTR::TestController::platformRunUntil): Use the new runRunLoopUntil function, and also use
447         it (with a timeout) when waiting for a crash log for the web process to be saved. This will
448         prevent us from waiting forever.
449
450 2011-03-01  Adam Roben  <aroben@apple.com>
451
452         Give WebKitLauncherWin's PDB file a name different from WebKit's
453
454         WebKitLauncherWin (which builds WebKit.exe) and WebKit (which builds WebKit.dll) were
455         fighting over the same PDB file, meaning that one of them was going to lose and not have
456         symbols. WebKitLauncherWin's PDB file is now named WebKitLauncherWin.pdb.
457
458         Fixes <http://webkit.org/b/55495> Crashes in WebKit.dll have bad backtraces for some test
459         runs (due to missing symbols)
460
461         Reviewed by David Kilzer.
462
463         * WebKitLauncherWin/WebKitLauncherWinCommon.vsprops: Use the project name,
464         "WebKitLauncherWin", instead of the target name, "WebKit", when naming the PDB file.
465
466 2011-03-01  Adam Roben  <aroben@apple.com>
467
468         Count web process crashes toward the --exit-after-n-crashes-or-timeouts limit
469
470         Fixes <http://webkit.org/b/55492> Web process crashes aren't counted toward the
471         --exit-after-n-crashes-or-timeouts limit
472
473         Reviewed by David Kilzer.
474
475         * Scripts/old-run-webkit-tests:
476         (stopRunningTestsEarlyIfNeeded): Include web process crashes in the computation, and print
477         the number of tests that caused the web process to crash when exiting early.
478
479 2011-03-01  Adam Roben  <aroben@apple.com>
480
481         Save a crash log when the web process crashes
482
483         On Windows, WebKitTestRunner now detects when the web process is crashing and waits to exit
484         until it has finished crashing, which guarantees that the crash log will have had time to be
485         saved, too. On Mac, we always wait until ReportCrash has exited before capturing the crash
486         log, so all we have to do is choose the right crash log out of the CrashReporter directory.
487
488         Fixes <http://webkit.org/b/44121> <rdar://problem/8320759> When the web process crashes and
489         a crash log is being saved, WebKitTestRunner thinks the web process has become unresponsive
490
491         Reviewed by Sam Weinig.
492
493         * Scripts/old-run-webkit-tests:
494         (testCrashedOrTimedOut): Don't kill WebKitTestRunner when the web process crashes. It will
495         kill itself. On Windows, this will cause us to wait until the crash log has been saved. On
496         Mac, it should have no effect. Capture saved crash logs for web process crashes, too.
497         (captureSavedCrashLog): Added $webProcessCrashed argument. On Mac, look for
498         WebProces_*.crash files when the web process crashes.
499
500         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
501         (WTR::InjectedBundle::initialize): Added an initializationUserData argument. Updated for
502         initializePlatformDefaults -> platformInitialize rename. Pass the initializationUserData
503         along to platformInitialize.
504
505         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: See above.
506
507         * WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
508         (WKBundleInitialize): Pass along the initializationUserData to the InjectedBundle.
509
510         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
511         (WTR::InjectedBundle::platformInitialize):
512         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
513         (WTR::InjectedBundle::platformInitialize):
514         Updated function signature.
515
516         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
517         (WTR::exceptionFilter): Added. Tells the UI process we're crashing by signaling the
518         webProcessCrashingEvent, then lets the crash continue as normal.
519
520         (WTR::InjectedBundle::platformInitialize): Hook up exceptionFilter. Retrieve the name of the
521         event we should use to tell the UI process we're crashing from the initializationUserData,
522         and get a handle to that event.
523
524         * WebKitTestRunner/TestController.cpp:
525         (WTR::TestController::TestController): Initialize new members.
526         (WTR::TestController::processDidCrash): Removed unnecessary WKPageRef argument. Changed to
527         only print the "#CRASHED - WebProcess" message once, since this can be called more than once
528         when a crash log is being saved on Windows. Exit right away if specified. (This is the
529         default.)
530
531         * WebKitTestRunner/TestController.h: Added new members.
532
533         * WebKitTestRunner/win/TestControllerWin.cpp:
534         (WTR::TestController::platformInitialize): Set up the event the web process will use to tell
535         us it's crashing.
536         (WTR::TestController::platformRunUntil): Pass MWMO_INPUTAVAILABLE to
537         ::MsgWaitForMultipleObjectsEx so we'll process messages that have already been seen by
538         ::PeekMessage. (This is unrelated to the bug fix.) Notice when the webProcessCrashingEvent
539         has been signaled. When this happens, print the "#CRASHED - WebProcess" message right away
540         so the test harness will know the web process has crashed and not try to kill us, then wait
541         for the web process to finish crashing so a crash log will have time to be saved.
542         (WTR::toWK): Simple hepler function.
543         (WTR::TestController::platformInitializeContext): Pass along the name of the event the web
544         process should use to tell us it is crashing in the context's initialization user data.
545
546 2011-03-01  Dimitri Glazkov  <dglazkov@chromium.org>
547
548         Reviewed by Tony Gentilcore.
549
550         Move the checks to exclude Python/Perl tests for Chromium Win build to the right place.
551         https://bugs.webkit.org/show_bug.cgi?id=55476
552
553         Chromium Win uses BuildAndTestFactory, not TestFactory.
554
555         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Moved to the right place.
556
557 2011-03-01  Adam Roben  <aroben@apple.com>
558
559         Test that the WebKit2 responsiveness timer doesn't fire too early
560
561         Test for <http://webkit.org/b/55417> <rdar://problem/9065287> RunLoop::Timer fires 1000x too
562         early on Windows
563
564         Reviewed by Anders Carlsson.
565
566         * TestWebKitAPI/PlatformUtilities.h: Added sleep().
567
568         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp: Added.
569         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
570         (TestWebKitAPI::didFinishLoadForFrame):
571         (TestWebKitAPI::processDidBecomeUnresponsive):
572         (TestWebKitAPI::setInjectedBundleClient):
573         (TestWebKitAPI::setPageLoaderClient):
574         Simple helper functions.
575
576         (TestWebKitAPI::TEST): Load an HTML file to make sure the web process is initialized. Then
577         tell the web process to pause and press the spacebar key. The spacebar keypress should cause
578         the responsiveness timer to start, but the web process should unpause before it has a chance
579         to fire. Run until the web process has unpaused, and assert that the timer didn't fire.
580
581         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly_Bundle.cpp: Added.
582         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::ResponsivenessTimerDoesntFireEarlyTest):
583         Call up to the base class.
584         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::didReceiveMessage): When asked to
585         pause, sleep for 0.5 seconds, then send back a message saying we paused.
586
587         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
588         (TestWebKitAPI::Util::sleep): Added. Calls through to usleep.
589         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
590         (TestWebKitAPI::Util::sleep): Added. Calls through to ::Sleep.
591
592         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
593         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
594         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
595         Added new files.
596
597 2011-03-01  Andras Becsi  <abecsi@webkit.org>
598
599         Reviewed by Csaba Osztrogonác.
600
601         [Qt] Clean up the project files and move common options to WebKit.pri.
602
603         * WebKitTestRunner/qt/WebKitTestRunner.pro: Deduplicate.
604
605
606 2011-03-01  Adam Roben  <aroben@apple.com>
607
608         Skip another multiprocessing test on Windows
609
610         It is sometimes failing (and may be leaving child Python processes around).
611         <http://webkit.org/b/55087> tracks the failure.
612
613         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
614         (FuncitonTests.test_get__processes): Skip this test on Windows.
615
616 2011-03-01  Ilya Tikhonovsky  <loislo@chromium.org>
617
618         Reviewed by Pavel Feldman.
619
620         Web Inspector: reduce inspector tests flakiness.
621         https://bugs.webkit.org/show_bug.cgi?id=55401
622
623         * DumpRenderTree/chromium/TestShell.cpp:
624         (TestShell::runFileTest):
625         * DumpRenderTree/chromium/WebPreferences.cpp:
626         (WebPreferences::reset):
627
628 2011-02-28  Mihai Parparita  <mihaip@chromium.org>
629
630         Reviewed by James Robinson.
631
632         NRWT should still upload results when it exits early due to too many failures
633         https://bugs.webkit.org/show_bug.cgi?id=55432
634         
635         Only keyboard interrupts should prevent results from being uploaded.
636
637         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
638         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
639
640 2011-02-28  Alice Liu  <alice.liu@apple.com>
641
642         Reviewed by Adam Roben and Gavin Barraclough.
643
644         Test for https://bugs.webkit.org/show_bug.cgi?id=54898
645         InjectedBundleNodeHandle dies too early in WKBundleHitTestResultGetNodeHandle.
646
647         * TestWebKitAPI/InjectedBundleController.h:
648         (TestWebKitAPI::InjectedBundleController::bundle):
649         * TestWebKitAPI/PlatformWebView.h:
650         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp: Added.
651         (TestWebKitAPI::didFinishLoadForFrame):
652         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
653         (TestWebKitAPI::setPageLoaderClient):
654         (TestWebKitAPI::TEST):
655         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp: Added.
656         (TestWebKitAPI::HitTestResultNodeHandleTest::HitTestResultNodeHandleTest):
657         (TestWebKitAPI::HitTestResultNodeHandleTest::getContextMenuFromDefaultMenu):
658         (TestWebKitAPI::HitTestResultNodeHandleTest::didCreatePage):
659         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
660         (TestWebKitAPI::PlatformWebView::simulateRightClick):
661         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
662         (TestWebKitAPI::PlatformWebView::simulateRightClick):
663         
664         Added files to:
665         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
666         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
667         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
668
669 2011-02-28  Adam Roben  <aroben@apple.com>
670
671         Start running InjectedBundle API tests on Windows
672
673         A new project, TestWebKitAPIInjectedBundle, builds the injected bundle DLL.
674
675         Fixes <http://webkit.org/b/55420> run-api-tests doesn't run injected bundle tests on Windows
676
677         Reviewed by Darin Adler.
678
679         * TestWebKitAPI/Configurations/TestWebKitAPIDebug.vsprops: Added.
680         * TestWebKitAPI/Configurations/TestWebKitAPIDebugAll.vsprops: Added.
681         * TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops: Added.
682         * TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops: Added.
683         * TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops: Added.
684         These are just extracted from TestWebKitAPI.vcproj.
685
686         * TestWebKitAPI/Configurations/TestWebKitAPIInjectedBundleCommon.vsprops: Added. Just sets
687         up the DLL's filename.
688
689         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Fixed DLL name.
690         (TestWebKitAPI::Util::createInjectedBundlePath): Changed to use kCFAllocatorNull instead of
691         0 (== kCFAllocatorDefault) so that CF won't try to deallocate our C string.
692
693         * TestWebKitAPI/win/TestWebKitAPI.sln: Added TestWebKitAPIInjectedBundle. Removed the bogus
694         "all" configuration.
695
696         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Changed to use the new .vsprops files. Added
697         injected bundle tests. Let VS resort things.
698
699         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Copied from Tools/TestWebKitAPI/win/TestWebKitAPI.vcproj.
700
701 2011-02-28  Adam Roben  <aroben@apple.com>
702
703         Add a test for the paint rect passed via WM_PAINT to windowless plugins
704
705         Test for <http://webkit.org/b/55365> <rdar://problem/9031089> REGRESSION (r79040): WebKit2:
706         Large portions of pages with plugins paint black
707
708         Reviewed by Anders Carlsson.
709
710         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
711         (PluginTest::NPP_HandleEvent):
712         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
713         Added NPP_HandleEvent. Just returns 0 at this level.
714
715         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowlessPaintRectCoordinates.cpp: Added.
716         (WindowlessPaintRectCoordinates::WindowlessPaintRectCoordinates): Call up to the base class.
717         (WindowlessPaintRectCoordinates::NPP_New): Mark ourselves as windowless.
718         (WindowlessPaintRectCoordinates::NPP_HandleEvent): Check that the paint rect passed via
719         WM_PAINT has the expected coordinates.
720
721         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
722         (NPP_HandleEvent): Give the PluginTest a chance to handle the event before doing anything
723         else.
724
725         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new file, let VS
726         resort things.
727
728         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
729         (WTR::InjectedBundlePage::didFinishLoadForFrame): Added a call to WKBundlePageForceRepaint.
730         This is roughly equivalent to the call to -[WebView displayIfNeeded] in -[FrameLoadDelegate
731         webView:didFinishLoadForFrame:] in DumpRenderTree.
732
733         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
734         (WTR::PlatformWebView::PlatformWebView): Tell the WKView it's in a window so that plugins
735         can start running.
736
737 2011-02-28  Adam Roben  <aroben@apple.com>
738
739         Tell MiniBrowser's WKViews that they're in a window
740
741         Fixes <http://webkit.org/b/55364> Plugins don't work in MiniBrowser on Windows
742
743         Reviewed by Anders Carlsson.
744
745         * MiniBrowser/win/BrowserView.cpp:
746         (BrowserView::create): Call WKViewSetIsInWindow after creating the view.
747
748 2011-02-28  Alexis Menard  <alexis.menard@openbossa.org>
749
750         Reviewed by Csaba Osztrogonác.
751
752         The new QML tests require QML_IMPORT_PATH variable to be set in order to
753         load the proper plugin.
754
755         * Scripts/run-qtwebkit-tests:
756
757 2011-02-28  Joone Hur  <joone.hur@collabora.co.uk>
758
759         Unreviewed, fix spelling mistake.
760
761         hardware-accelated => hardware-accelerated.
762
763         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
764
765 2011-02-28  Kristian Amlie  <kristian.amlie@nokia.com>
766
767         Reviewed by Andreas Kling.
768
769         Made sure that the build-webkit --qmake argument is always respected.
770
771         Previously, it would be respected during the build, but not during
772         the early feature detection. The build would also fail if qmake was
773         not in your path, even if you specified --qmake with a valid qmake
774         binary.
775
776         [Qt] WebKit patches required to work with a modularized version of Qt
777         https://bugs.webkit.org/show_bug.cgi?id=53916
778
779         * Scripts/build-webkit:
780         * Scripts/webkitdirs.pm:
781
782 2011-02-27  Ojan Vafai  <ojan@chromium.org>
783
784         Fix failing tests from http://trac.webkit.org/changeset/79837.
785
786         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
787         This was just a mistake that would break if someone tried printing with this configuration.
788         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
789         Method signature changed. Didn't notice that this method was being calling in a unittest.
790
791 2011-02-24  Hayato Ito  <hayato@chromium.org>
792
793         Reviewed by Eric Seidel.
794
795         [NRWT] Remove test_types/* classes.
796
797         This is the last patch in the series of efforts which tried to remove test_types classes.
798         A single_test_runner module will take over responsibilities of test_type classes.
799
800         From this patch, we won't display statistics per test_type even if a '--verbose' option is given.
801
802         https://bugs.webkit.org/show_bug.cgi?id=55123
803
804         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
805         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
806         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
807         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
808         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
809         * Scripts/webkitpy/layout_tests/test_types/__init__.py: Removed.
810         * Scripts/webkitpy/layout_tests/test_types/image_diff.py: Removed.
811         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py: Removed.
812         * Scripts/webkitpy/layout_tests/test_types/test_type_base_unittest.py: Removed.
813         * Scripts/webkitpy/layout_tests/test_types/text_diff.py: Removed.
814
815 2011-02-25  Ojan Vafai  <ojan@chromium.org>
816
817         Reviewed by Tony Chang.
818
819         Change results.json format to the one used by unexpected_results.json
820         https://bugs.webkit.org/show_bug.cgi?id=52267
821
822         Also add runtimes in milliseconds to the JSON and make the output format more compact.
823         Named the file full_results.json to avoid conflicting with the results.json
824         file the test-results server currently serves up.
825
826         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
827         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
828         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
829         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
830         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
831         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
832
833 2011-02-27  Adam Roben  <aroben@apple.com>
834
835         Followup fix to r79827
836
837         * DumpRenderTree/win/UIDelegate.cpp:
838         (UIDelegate::webViewAddMessageToConsole): Replaced call to urlSuitableForTestResult with
839         lastPathComponent. This matches what we did here before r79827, and matches Mac.
840
841 2011-02-27  Dan Bernstein  <mitz@apple.com>
842
843         Reviewed by Adam Roben.
844
845         Make bisect-builds print a trac link for the regression range
846         https://bugs.webkit.org/show_bug.cgi?id=55331
847
848         * Scripts/bisect-builds:
849
850 2011-02-27  Adam Roben  <aroben@apple.com>
851
852         Use iswalpha instead of isalpha when dealing with wchar_ts in EditingDelegate
853
854         Fixes <http://webkit.org/b/55062> <rdar://problem/9059907> Crash beneath
855         EditingDelegate::checkSpellingOfString when running fast/forms/input-text-maxlength.html or
856         fast/forms/input-text-paste-maxlength.html on Windows with full page heap enabled
857
858         Reviewed by Anders Carlsson.
859
860         * DumpRenderTree/win/EditingDelegate.cpp:
861         (indexOfFirstWordCharacter):
862         (wordLength):
863         Use iswalpha instead of isalpha, since TCHAR is really just wchar_t.
864
865 2011-02-27  Adam Roben  <aroben@apple.com>
866
867         Make ResourceLoadDelegate print URLs relative to the main resource on Windows
868
869         This matches Mac. (In fact, the code was ported from Mac!)
870
871         Fixes <http://webkit.org/b/55328> URLs printed by ResourceLoadDelegate on Windows don't
872         match Mac
873
874         Reviewed by Anders Carlsson.
875
876         * DumpRenderTree/win/DumpRenderTree.cpp:
877         (substringFromIndex): Added. Emulates -[NSString substringFromIndex:].
878         (urlSuitableForTestResult): Ported code from -[NSURL(DRTExtras)
879         _drt_descriptionSuitableForTestResult].
880         (cfStringRefToWString): Moved here from LayoutTestControllerWin.cpp.
881
882         * DumpRenderTree/win/DumpRenderTreeWin.h: Added declaration of cfStringRefToWString.
883
884         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Moved cfStringRefToWString to
885         DumpRenderTree.cpp.
886
887 2011-02-27  Adam Roben  <aroben@apple.com>
888
889         Change the order of Chang Shu's email addresses so the first one is their Bugzilla username
890
891         This makes Bugzilla autocompletion work correctly.
892
893         * Scripts/webkitpy/common/config/committers.py:
894
895 2011-02-27  Patrick Gansterer  <paroga@webkit.org>
896
897         Reviewed by Andreas Kling.
898
899         [Qt] fast/loader/user-stylesheet-fast-path.html fails
900         https://bugs.webkit.org/show_bug.cgi?id=50911
901
902         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
903         (LayoutTestController::setUserStyleSheetLocation):
904         Use QUrl::fromEncoded since the passed string is already encoded.
905
906 2011-02-27  Adam Roben  <aroben@apple.com>
907
908         Ensure $testResultsDirectory is an absolute path before setting up Windows crash log saving
909
910         Windows 7 (and maybe Vista?) launches the post-mortem debugger with a working directory that
911         is different from old-run-webkit-test's. Because of this, the path to the crash log file
912         (which is relative to $testResultsDirectory) must be an absolute path. See the bug for more
913         details.
914
915         Fixes <http://webkit.org/b/55318> Crash logs aren't saved on Windows 7 when a relative
916         --results-directory path is used (like on the test slaves)
917
918         Reviewed by Dan Bernstein.
919
920         * Scripts/old-run-webkit-tests: Moved the call to setUpWindowsCrashLogSaving after we've
921         made $testResultsDirectory an absolute path.
922
923 2011-02-26  Dominic Mazzoni  <dmazzoni@google.com>
924
925         Reviewed by Adam Barth.
926
927         m_dumpAccessibilityNotifications should be initialized.
928         https://bugs.webkit.org/show_bug.cgi?id=54874
929
930         * DumpRenderTree/chromium/AccessibilityController.cpp:
931         (AccessibilityController::AccessibilityController):
932
933 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
934
935         Reviewed by Pavel Feldman.
936
937         DumpRenderTree should reset frame opener between tests.
938         https://bugs.webkit.org/show_bug.cgi?id=54874
939
940         Added clearOpener() calls between tests.
941         Some tests are using window.open() to change current window location.
942         This changes window.opener property whose nullity is checked before
943         showing xml viewer.
944
945         * DumpRenderTree/chromium/TestShell.cpp:
946         (TestShell::resetTestController):
947         * DumpRenderTree/gtk/DumpRenderTree.cpp:
948         (resetDefaultsToConsistentValues):
949         * DumpRenderTree/mac/DumpRenderTree.mm:
950         (resetWebViewToConsistentStateBeforeTesting):
951         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
952         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
953         * DumpRenderTree/win/DumpRenderTree.cpp:
954         (resetWebViewToConsistentStateBeforeTesting):
955         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
956         (WTR::InjectedBundlePage::reset):
957
958 2011-02-26  Yael Aharon  <yael.aharon@nokia.com>
959
960         Reviewed by Andreas Kling.
961
962         [Qt] Fix the focus in MiniBrowser.
963         https://bugs.webkit.org/show_bug.cgi?id=55288
964
965         Give focus to the web view when a load starts.
966
967         * MiniBrowser/qt/BrowserWindow.cpp:
968         (BrowserWindow::urlChanged):
969
970 2011-02-26  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
971
972         Reviewed by Adam Barth.
973
974         [EFL] Add coding style exceptions for EFL port
975         https://bugs.webkit.org/show_bug.cgi?id=54733
976
977         The EFL APIs use EFL naming style, which includes both lower-cased and camel-cased,
978         underscore-sparated values. But, style checking rule have found parameter_name errors
979         in efl directories so far.
980
981         * Scripts/webkitpy/style/checker.py: Add prameter_name rule's exception and efl directories.
982
983 2011-02-26  David Levin  <levin@chromium.org>
984
985         Reviewed by Darin Adler.
986
987         check-webkit-style should check for spaces after periods in comments
988         https://bugs.webkit.org/show_bug.cgi?id=55253
989
990         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
991         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests
992           and fixed the name of one related test.
993
994 2011-02-26  Kenneth Rohde Christiansen  <kenneth@webkit.org>
995
996         Reviewed by Andreas Kling.
997
998         Make it possible to test the targetdensity-dpi support
999         https://bugs.webkit.org/show_bug.cgi?id=55142
1000
1001         Test the viewport meta tag feature targetdensity-dpi by
1002         adding extra arguments to dumpConfigurationForViewport
1003
1004         Add the new arguments to the shared header and all
1005         implementation.
1006
1007         * DumpRenderTree/LayoutTestController.cpp:
1008         (dumpConfigurationForViewportCallback):
1009         * DumpRenderTree/LayoutTestController.h:
1010         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1011         (LayoutTestController::dumpConfigurationForViewport):
1012         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1013         (LayoutTestController::dumpConfigurationForViewport):
1014         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1015         (LayoutTestController::dumpConfigurationForViewport):
1016         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1017         (LayoutTestController::dumpConfigurationForViewport):
1018         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1019         (LayoutTestController::dumpConfigurationForViewport):
1020         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1021
1022 2011-02-26  Sheriff Bot  <webkit.review.bot@gmail.com>
1023
1024         Unreviewed, rolling out r79764.
1025         http://trac.webkit.org/changeset/79764
1026         https://bugs.webkit.org/show_bug.cgi?id=55295
1027
1028         "broke Chromium builds" (Requested by rniwa on #webkit).
1029
1030         * DumpRenderTree/chromium/TestShell.cpp:
1031         (TestShell::resetTestController):
1032         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1033         (resetDefaultsToConsistentValues):
1034         * DumpRenderTree/mac/DumpRenderTree.mm:
1035         (resetWebViewToConsistentStateBeforeTesting):
1036         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1037         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1038         * DumpRenderTree/win/DumpRenderTree.cpp:
1039         (resetWebViewToConsistentStateBeforeTesting):
1040         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1041         (WTR::InjectedBundlePage::reset):
1042
1043 2011-02-26  Yongjun Zhang  <yongjun_zhang@apple.com>
1044
1045         Reviewed by David Kilzer.
1046
1047         https://bugs.webkit.org/show_bug.cgi?id=48781
1048
1049         Add a resource load delegate method to query if WebCore should paint the default broken image for failed images.
1050
1051         Add a new resource load client method (shouldPaintBrokenImage).  WebKit client can decide if WebCore
1052         should paint the default broken image when an image fails to load or decode.
1053
1054         * DumpRenderTree/LayoutTestController.cpp:
1055         (LayoutTestController::LayoutTestController):
1056         (setShouldPaintBrokenImageCallback):
1057         (LayoutTestController::staticFunctions):
1058         (LayoutTestController::setShouldPaintBrokenImage):
1059         * DumpRenderTree/LayoutTestController.h:
1060         (LayoutTestController::shouldPaintBrokenImage):
1061         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
1062         (-[ResourceLoadDelegate webView:shouldPaintBrokenImageForURL:]):
1063
1064 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
1065
1066         Reviewed by Pavel Feldman.
1067
1068         DumpRenderTree should reset frame opener between tests.
1069         https://bugs.webkit.org/show_bug.cgi?id=54874
1070
1071         Added clearOpener() calls between tests.
1072         Some tests are using window.open() to change current window location.
1073         This changes window.opener property whose nullity is checked before
1074         showing xml viewer.
1075
1076         * DumpRenderTree/chromium/TestShell.cpp:
1077         (TestShell::resetTestController):
1078         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1079         (resetDefaultsToConsistentValues):
1080         * DumpRenderTree/mac/DumpRenderTree.mm:
1081         (resetWebViewToConsistentStateBeforeTesting):
1082         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1083         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1084         * DumpRenderTree/win/DumpRenderTree.cpp:
1085         (resetWebViewToConsistentStateBeforeTesting):
1086         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1087         (WTR::InjectedBundlePage::reset):
1088
1089 2011-02-25  Lukasz Slachciak  <l.slachciak@samsung.com>
1090
1091         Reviewed by Kenneth Rohde Christiansen.
1092
1093         [EFL] Added EWebLauncher command line option to turn on/off frame flattening
1094         https://bugs.webkit.org/show_bug.cgi?id=54809
1095
1096         * EWebLauncher/main.c: Handling -f option added.
1097         (on_key_down):
1098         (browserCreate):
1099         (main):
1100
1101 2011-02-25  Sheriff Bot  <webkit.review.bot@gmail.com>
1102
1103         Unreviewed, rolling out r79324.
1104         http://trac.webkit.org/changeset/79324
1105         https://bugs.webkit.org/show_bug.cgi?id=55260
1106
1107         Leopard Debug is once again running out of address space while
1108         linking (Requested by aroben on #webkit).
1109
1110         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1111         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1112
1113 2011-02-25  Tony Chang  <tony@chromium.org>
1114
1115         Reviewed by James Robinson.
1116
1117         Clean up temp files left by crashing DRTs on chromium-win
1118         https://bugs.webkit.org/show_bug.cgi?id=55182
1119
1120         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1121         * BuildSlaveSupport/chromium/remove-crash-logs: Added.
1122
1123 2011-02-25  Adam Roben  <aroben@apple.com>
1124
1125         Double-check that modifying the registry worked
1126
1127         On Windows Vista/7 with UAC enabled, regtool will fail to modify the registry, but it will
1128         still return a successful exit code. So we double-check here that the value we tried to
1129         write to the registry was really written.
1130
1131         Fixes <http://webkit.org/b/55227> old-run-webkit-tests falsely claims to be saving crash
1132         logs on Windows Vista/7
1133
1134         Reviewed by Darin Adler.
1135
1136         * Scripts/old-run-webkit-tests:
1137         (writeRegistryString): After trying to write the string value, check that it was really
1138         written.
1139
1140 2011-02-25  Adam Roben  <aroben@apple.com>
1141
1142         Set the 32-bit post-mortem debugger on 64-bit OSes
1143
1144         The post-mortem debugger is controlled by some registry values in
1145         HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AeDebug. On 64-bit OSes, you can either
1146         set the values in that key, which will be used for 64-bit apps, or in the equivalent key
1147         under Wow6432Node, which will be used for 32-bit apps. Since DumpRenderTree and
1148         WebKitTestRunner are 32-bit, we want to use the latter location.
1149
1150         Fixes <http://webkit.org/b/55225> old-run-webkit-tests sets the post-mortem debugger for
1151         64-bit apps, but should set it for 32-bit apps instead
1152
1153         Reviewed by Darin Adler.
1154
1155         * Scripts/old-run-webkit-tests:
1156         (readRegistryString):
1157         (writeRegistryString):
1158         Pass --wow32 to regtool so it will get/set values beneath Wow6432Node, which is where
1159         Windows looks for the post-mortem debugger values when a 32-bit app crashes on a 64-bit OS.
1160         Passing --wow32 on a 32-bit OS will not have any effect.
1161
1162 2011-02-25  Adam Roben  <aroben@apple.com>
1163
1164         Clean up code to set up crash log saving on Windows
1165
1166         Fixes <http://webkit.org/b/55224> old-run-webkit-test's Windows registry code is ugly
1167
1168         Reviewed by David Kilzer.
1169
1170         * Scripts/old-run-webkit-tests:
1171         (readRegistryString):
1172         (writeRegistryString):
1173         Moved code to read strings from and write strings to the registry here...
1174
1175         (setUpWindowsCrashLogSaving): ...from here.
1176
1177 2011-02-25  Adam Roben  <aroben@apple.com>
1178
1179         Work around Cygwin's crash-suppression behavior
1180
1181         Cygwin calls ::SetErrorMode(SEM_FAILCRITICALERRORS), which any processes it launches will
1182         inherit. This is bad for testing/debugging, as it causes the post-mortem debugger not to be
1183         invoked. (Cygwin does this because it makes crashes more UNIX-y.) We reset the error mode
1184         when our test apps launch to work around Cygwin's behavior.
1185
1186         Fixes <http://webkit.org/b/55222> Test apps crash silently (without invoking post-mortem
1187         debugger) when launched from Cygwin 1.7
1188
1189         Reviewed by Darin Adler.
1190
1191         * DumpRenderTree/win/DumpRenderTree.cpp:
1192         (main):
1193         * TestWebKitAPI/win/main.cpp:
1194         (main):
1195         * WebKitAPITest/main.cpp:
1196         (main):
1197         * WebKitTestRunner/win/TestControllerWin.cpp:
1198         (WTR::TestController::platformInitialize):
1199         Call ::SetErrorMode(0) to undo Cygwin's folly.
1200
1201 2011-02-24  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
1202
1203         Reviewed by Andreas Kling.
1204
1205         [Qt] Revert the support for QNAM affined to a different thread.
1206         https://bugs.webkit.org/show_bug.cgi?id=55149
1207
1208         Qt 4.8 will have QNAM use its own thread internally by default,
1209         no need to keep this complexity in WebKit.
1210
1211         This mainly reverts:
1212         http://trac.webkit.org/changeset/73710
1213         http://trac.webkit.org/changeset/73712
1214
1215         * QtTestBrowser/launcherwindow.cpp:
1216         (LauncherWindow::~LauncherWindow):
1217         (LauncherWindow::initializeView):
1218         (LauncherWindow::createChrome):
1219         * QtTestBrowser/launcherwindow.h:
1220         (WindowOptions::WindowOptions):
1221         * QtTestBrowser/webpage.cpp:
1222         (WebPage::WebPage):
1223         * QtTestBrowser/webpage.h:
1224
1225 2011-02-24  Adam Barth  <abarth@webkit.org>
1226
1227         Reviewed by Eric Seidel.
1228
1229         Add a script for creating gypi files
1230         https://bugs.webkit.org/show_bug.cgi?id=55113
1231
1232         This is a trivial script to create gypi files from the source code in
1233         the current directory.
1234
1235         * Scripts/make-gypi: Added.
1236
1237 2011-02-24  Sheriff Bot  <webkit.review.bot@gmail.com>
1238
1239         Unreviewed, rolling out r79628.
1240         http://trac.webkit.org/changeset/79628
1241         https://bugs.webkit.org/show_bug.cgi?id=55195
1242
1243         "broke chromium win tests" (Requested by dpranke on #webkit).
1244
1245         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1246         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1247
1248 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
1249
1250         Reviewed by Mihai Parparita.
1251
1252         rebaseline-chromium-webkit-tests can't handle GPU variants
1253         https://bugs.webkit.org/show_bug.cgi?id=55002
1254
1255         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1256         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
1257
1258 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
1259
1260         Reviewed by Mihai Parparita.
1261
1262         nrwt: log command used to start web server to --verbose
1263
1264         https://bugs.webkit.org/show_bug.cgi?id=55188
1265
1266         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1267         * Scripts/webkitpy/layout_tests/port/http_server.py:
1268
1269 2011-02-24  Victoria Kirst  <vrk@google.com>
1270
1271         Reviewed by Mihai Parparita.
1272
1273         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
1274         https://bugs.webkit.org/show_bug.cgi?id=55089
1275
1276         This adds an alias into the media/ directory so that http/tests/media
1277         tests can access the media resources when running in an httpd process.
1278
1279         * Scripts/webkitperl/httpd.pm:
1280         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1281         * Scripts/webkitpy/layout_tests/port/http_server.py:
1282
1283 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
1284
1285         Reviewed by Mihai Parparita.
1286
1287         nrwt: enable multiple processes by default on Chromium Win.
1288
1289         https://bugs.webkit.org/show_bug.cgi?id=55163
1290
1291         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1292         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1293
1294 2011-02-24  Sam Weinig  <sam@webkit.org>
1295
1296         Attempt to fix Qt.
1297
1298         * WebKitTestRunner/TestInvocation.cpp:
1299
1300 2011-02-24  James Robinson  <jamesr@chromium.org>
1301
1302         Unreviewed, rolling out r79598.
1303         http://trac.webkit.org/changeset/79598
1304         https://bugs.webkit.org/show_bug.cgi?id=55089
1305
1306         Appears to be causing trouble with httpd on windows
1307
1308         * Scripts/webkitperl/httpd.pm:
1309         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1310         * Scripts/webkitpy/layout_tests/port/http_server.py:
1311
1312 2011-02-24  Sam Weinig  <sam@webkit.org>
1313
1314         Reviewed by Darin Adler.
1315
1316         Add basic pixel testing support to WebKitTestRunner
1317         https://bugs.webkit.org/show_bug.cgi?id=55161
1318
1319         * Scripts/old-run-webkit-tests:
1320         Don't print color space information when using WebKitTestRunner since
1321         we don't need to change the screen.
1322
1323         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1324         (WTR::InjectedBundle::initialize):
1325         Add call to initializePlatformDefaults.
1326         
1327         (WTR::InjectedBundle::done):
1328         Return a dictionary on completion with both the text output and pixel dump override.
1329
1330         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1331         Add declaration for initializePlatformDefaults.
1332
1333         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1334         (WTR::LayoutTestController::LayoutTestController):
1335         (WTR::LayoutTestController::dumpAsText):
1336         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1337         (WTR::LayoutTestController::shouldDumpPixels):
1338         Make setting dumpAsText override the dump pixels preference.
1339
1340         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: Added.
1341         (WTR::InjectedBundle::initializePlatformDefaults):
1342         Set platform defaults matching DumpRenderTree.
1343
1344         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp: Added.
1345         (WTR::InjectedBundle::initializePlatformDefaults):
1346         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp: Added.
1347         (WTR::InjectedBundle::initializePlatformDefaults):
1348         Add stubs for initializePlatformDefaults.
1349
1350         * WebKitTestRunner/TestController.cpp:
1351         (WTR::TestController::runTest):
1352         (WTR::TestController::runTestingServerLoop):
1353         (WTR::TestController::run):
1354         Parse expected pixel results out of stdin and setup the invocation with it.
1355
1356         * WebKitTestRunner/TestInvocation.cpp:
1357         (WTR::TestInvocation::TestInvocation):
1358         Initialize new members.
1359
1360         (WTR::TestInvocation::~TestInvocation):
1361         (WTR::TestInvocation::setIsPixelTest):
1362         (WTR::sizeWebViewForCurrentTest):
1363         (WTR::TestInvocation::invoke):
1364         (WTR::TestInvocation::dump):
1365         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
1366         * WebKitTestRunner/TestInvocation.h:
1367         Dump pixels in addition to text by calling dumpPixelsAndCompareWithExpected.
1368
1369         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
1370         (WTR::PlatformWebView::PlatformWebView):
1371         Use -[NSColorSpace genericRGBColorSpace] instead of the main screen color space.
1372
1373         * WebKitTestRunner/mac/TestInvocationMac.mm: Added.
1374         (WTR::createCGContextFromPlatformView):
1375         (WTR::computeMD5HashStringForContext):
1376         (WTR::dumpBitmap):
1377         (WTR::forceRepaintFunction):
1378         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
1379         Add mac specific pixel dumping support. Right now, we always pull pixels
1380         from the window server.
1381
1382         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
1383         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
1384         * WebKitTestRunner/qt/WebKitTestRunner.pro:
1385         * WebKitTestRunner/win/InjectedBundle.vcproj:
1386         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
1387         Add new files.
1388
1389 2011-02-24  Chang Shu  <cshu@webkit.org>
1390
1391         Reviewed by Kenneth Rohde Christiansen.
1392
1393         [Qt] Launch MiniBrowser if the test run is WebKit2. 
1394         https://bugs.webkit.org/show_bug.cgi?id=55145
1395
1396         * Scripts/old-run-webkit-tests:
1397         * Scripts/run-launcher:
1398         * Scripts/webkitdirs.pm:
1399
1400 2011-02-24  Victoria Kirst  <vrk@google.com>
1401
1402         Reviewed by Mihai Parparita.
1403
1404         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
1405         https://bugs.webkit.org/show_bug.cgi?id=55089
1406
1407         This adds an alias into the media/ directory so that http/tests/media
1408         tests can access the media resources when running in an httpd process.
1409
1410         * Scripts/webkitperl/httpd.pm:
1411         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1412         * Scripts/webkitpy/layout_tests/port/http_server.py:
1413
1414 2011-02-24  Andrew Wilson  <atwilson@chromium.org>
1415
1416         Unreviewed, rolling out r79570.
1417         http://trac.webkit.org/changeset/79570
1418         https://bugs.webkit.org/show_bug.cgi?id=54874
1419
1420         Breaks chromium build because glue/mocks/mock_web_frame.h/cc
1421         was not updated
1422
1423         * DumpRenderTree/chromium/TestShell.cpp:
1424         (TestShell::resetTestController):
1425         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1426         (resetDefaultsToConsistentValues):
1427         * DumpRenderTree/mac/DumpRenderTree.mm:
1428         (resetWebViewToConsistentStateBeforeTesting):
1429         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1430         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1431         * DumpRenderTree/win/DumpRenderTree.cpp:
1432         (resetWebViewToConsistentStateBeforeTesting):
1433         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1434         (WTR::InjectedBundlePage::reset):
1435
1436 2011-02-24  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
1437
1438         Reviewed by Laszlo Gombos.
1439
1440         [Qt] add android 2.2 user agent to QtTestBrowser
1441         https://bugs.webkit.org/show_bug.cgi?id=55085
1442
1443         Credit to Forrest Hodgkins to find right user-agent which works for youtube.
1444         * QtTestBrowser/useragentlist.txt:
1445
1446 2011-02-24  Adam Roben  <aroben@apple.com>
1447
1448         Change FrameLoadDelegate to support any number of delegates with delayed work to process
1449
1450         This makes our behavior match Mac more closely, and allows us to remove an incorrect
1451         assertion that was firing during some tests. (The assertion was claiming that there was
1452         never more than one delegate with delayed work to process, but that was not the case.)
1453
1454         Fixes <http://webkit.org/b/55146> Assertion failure in FrameLoadDelegate::locationChangeDone
1455         when running http/tests/navigation/back-twice-without-commit.html
1456
1457         Reviewed by Eric Carlson.
1458
1459         * DumpRenderTree/win/FrameLoadDelegate.cpp:
1460         (delegatesWithDelayedWork): Added. Returns all FrameLoadDelegates that have delayed work to
1461         process. A single delegate may appear in this Vector more than once (just as, on Mac, a
1462         single delegate may have multiple performSelector requests).
1463         (processWorkTimer): Pass the HWND to ::KillTimer, for pedantic brownie points. Added an
1464         assertion that the timer firing is the shared process work timer. Instead of using the
1465         single, global "delegate waiting for timer" delegate, give all delegates that have delayed
1466         work to process a chance to process their work.
1467         (FrameLoadDelegate::locationChangeDone): If we don't already have an active timer for
1468         processing delayed work, create one. Then add ourselves to the delegatesWithDelayedWork
1469         Vector so our processWork function will be called when the timer fires.
1470
1471 2011-02-24  Vsevolod Vlasov  <vsevik@chromium.org>
1472
1473         Reviewed by Alexey Proskuryakov.
1474
1475         DumpRenderTree should reset frame opener between tests.
1476         https://bugs.webkit.org/show_bug.cgi?id=54874
1477
1478         Added clearOpener() calls between tests.
1479         Some tests are using window.open() to change current window location.
1480         This changes window.opener property whose nullity is checked before
1481         showing xml viewer.
1482
1483         * DumpRenderTree/chromium/TestShell.cpp:
1484         (TestShell::resetTestController):
1485         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1486         (resetDefaultsToConsistentValues):
1487         * DumpRenderTree/mac/DumpRenderTree.mm:
1488         (resetWebViewToConsistentStateBeforeTesting):
1489         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1490         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1491         * DumpRenderTree/win/DumpRenderTree.cpp:
1492         (resetWebViewToConsistentStateBeforeTesting):
1493         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1494         (WTR::InjectedBundlePage::reset):
1495
1496 2011-02-24  Adam Roben  <aroben@apple.com>
1497
1498         Don't trigger Windows builds for changes to WebCore.exp.in or the top-level GNUmakefile.am
1499
1500         Fixes <http://webkit.org/b/55144> Windows builders should not have built r79343 or r79440,
1501         but did
1502
1503         Reviewed by Anders Carlsson.
1504
1505         * Scripts/webkitpy/common/config/build.py:
1506         (_should_file_trigger_build): Loosened the pattern for GNUmakefile.am to include the
1507         top-level GNUmakefile.am, and loosened the pattern for .exp files to include WebCore.exp.in.
1508
1509         * Scripts/webkitpy/common/config/build_unittest.py:
1510         (ShouldBuildTest): Added some more tests.
1511
1512 2011-02-24  Andras Becsi  <abecsi@webkit.org>
1513
1514         Reviewed by Laszlo Gombos.
1515
1516         [Qt] MinGW build fails to link
1517         https://bugs.webkit.org/show_bug.cgi?id=55050
1518
1519         Prepend the libraries of subcomponents instead of appending them
1520         to fix the library order according to the dependency of the libraries
1521
1522         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
1523         * WebKitTestRunner/qt/WebKitTestRunner.pro:
1524
1525 2011-02-24  Gabor Loki  <loki@webkit.org>
1526
1527         Reviewed by Csaba Osztrogonác.
1528
1529         [Qt] Remove ARMv5 Release buildbot.
1530
1531         Remove ARMv5 Release buildbot, because there is very little difference
1532         between ARMv5 and ARMv7 Release bots, and there is also very little
1533         interest in ARMv5 build slave (compared to ARMv7).
1534
1535         * BuildSlaveSupport/build.webkit.org-config/config.json:
1536         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1537
1538 2011-02-22  Ojan Vafai  <ojan@chromium.org>
1539
1540         Reviewed by Tony Chang.
1541
1542         fix revision number finding when creating git patches
1543         https://bugs.webkit.org/show_bug.cgi?id=55024
1544
1545         HEAD~n does not walk merge commits correctly. git log however does.
1546
1547         * Scripts/webkitpy/common/checkout/scm.py:
1548         * Scripts/webkitpy/common/checkout/scm_unittest.py:
1549
1550 2011-02-23  Kenneth Russell  <kbr@google.com>
1551
1552         Unreviewed, rolling out r79387.
1553         https://bugs.webkit.org/show_bug.cgi?id=54885
1554
1555         Reapply the code changes from r79038 since they weren't the root
1556         cause of the test timeouts.
1557
1558         * DumpRenderTree/chromium/WebViewHost.cpp:
1559         (WebViewHost::reset):
1560
1561 2011-02-23  Siddharth Mathur  <siddharth.mathur@nokia.com>
1562
1563         Reviewed by Laszlo Gombos.
1564
1565         [Qt] Make sure Symbian binary UIDs are unique
1566
1567         * MiniBrowser/qt/MiniBrowser.pro: Resolve the UID collision with
1568         QtTestBrowser.pro by changing the UID.
1569
1570 2011-02-23  Patrick Gansterer  <paroga@webkit.org>
1571
1572         Reviewed by Darin Adler.
1573
1574         Rename PLATFORM(CF) to USE(CF)
1575         https://bugs.webkit.org/show_bug.cgi?id=53540
1576
1577         * DumpRenderTree/config.h:
1578
1579 2011-02-23  Dirk Pranke  <dpranke@chromium.org>
1580
1581         Reviewed by Ojan Vafai.
1582
1583         nrwt: turn on multiprocessing on chromium_linux, clean up unit tests for worker model defaults
1584         https://bugs.webkit.org/show_bug.cgi?id=55016
1585
1586         * Scripts/webkitpy/layout_tests/port/base.py:
1587         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1588         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py: Added.
1589         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1590         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1591         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1592         * Scripts/webkitpy/layout_tests/port/mac.py:
1593         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1594         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
1595         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1596
1597 2011-02-23  Tony Chang  <tony@chromium.org>
1598
1599         Reviewed by Kent Tamura.
1600
1601         [chromium] DumpRenderTree --check-layout-test-sys-deps leaks 4 temp dirs
1602         https://bugs.webkit.org/show_bug.cgi?id=55004
1603
1604         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1605         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
1606         (WebKitSupportTestEnvironment::~WebKitSupportTestEnvironment):
1607         (main): Use a C++ object to ensure that TearDownTestEnvironment
1608             gets called when --check-layout-test-sys-deps is called
1609
1610 2011-02-23  Gopal Raghavan  <gopal.1.raghavan@nokia.com>
1611
1612         Reviewed by Anders Carlsson.
1613
1614         [Qt] [Qt] check webkit style fails in qt/tests
1615         https://bugs.webkit.org/show_bug.cgi?id=54998
1616
1617         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.
1618
1619         * Scripts/webkitpy/style/checker.py:
1620
1621 2011-02-22  Vsevolod Vlasov  <vsevik@chromium.org>
1622
1623         Reviewed by Pavel Feldman.
1624
1625         DumpRenderTree should dump as text if test path contains "dumpAsText/"
1626         https://bugs.webkit.org/show_bug.cgi?id=54873
1627
1628         * DumpRenderTree/chromium/LayoutTestController.h:
1629         (LayoutTestController::setShouldDumpAsText):
1630         (LayoutTestController::setShouldGeneratePixelResults):
1631         * DumpRenderTree/chromium/TestShell.cpp:
1632         (TestShell::runFileTest):
1633         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1634         (shouldDumpAsText):
1635         (dump):
1636         (runTest):
1637         * DumpRenderTree/mac/DumpRenderTree.mm:
1638         (shouldDumpAsText):
1639         (runTest):
1640         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1641         (WebCore::isDumpAsTextTest):
1642         (WebCore::DumpRenderTree::open):
1643         * DumpRenderTree/win/DumpRenderTree.cpp:
1644         (shouldDumpAsText):
1645         (runTest):
1646         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1647         (WTR::InjectedBundlePage::dump):
1648
1649 2011-02-22  Kenneth Russell  <kbr@google.com>
1650
1651         Reviewed by James Robinson.
1652
1653         Many tests started timing out on the chrome mac canaries around r79035
1654         https://bugs.webkit.org/show_bug.cgi?id=54885
1655
1656         Revert the code changes, but not the Chromium DEPS roll, from
1657         r79038 to see whether this clears up the timing-out tests. Was not
1658         able to reproduce the timeouts locally with a Release mode build
1659         of DRT (on Snow Leopard).
1660
1661         * DumpRenderTree/chromium/WebViewHost.cpp:
1662         (WebViewHost::reset):
1663
1664 2011-02-22  Dirk Pranke  <dpranke@chromium.org>
1665
1666         Reviewed by James Robinson.
1667
1668         nrwt: --platform chromium-gpu doesn't work with --worker-model=processes
1669         https://bugs.webkit.org/show_bug.cgi?id=55009
1670
1671         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1672
1673 2011-02-18  Mihai Parparita  <mihaip@chromium.org>
1674
1675         Reviewed by Ojan Vafai.
1676
1677         Default to 500 for --exit-after-n-failures
1678         https://bugs.webkit.org/show_bug.cgi?id=54773
1679
1680         Make 500 the default value for --exit-after-n-failures (ORWT bots
1681         already default to this as of r75726). Similarly, change the default for
1682         --exit-after-n-crashes-or-timeouts to 20.
1683
1684         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1685
1686 2011-02-02  Martin Robinson  <mrobinson@igalia.com>
1687
1688         Reviewed by Xan Lopez.
1689
1690         [GTK] DRT needs an implementation of LayoutTestController.setSelectTrailingWhitespace
1691         https://bugs.webkit.org/show_bug.cgi?id=53603
1692
1693         Add support for LayoutTestController.setSelectTrailingWhitespace by calling through
1694         to DumpRenderTreeSupportGtk for this functionality.
1695
1696         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1697         (resetDefaultsToConsistentValues): Always call LayoutTestController.setSelectTrailingWhitespace to
1698         false when reseting values between tests.
1699         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1700         (LayoutTestController::setSelectTrailingWhitespaceEnabled): Call LayoutTestController.setSelectTrailingWhitespace
1701         with the appropriate value.
1702
1703 2011-02-22  Adam Roben  <aroben@apple.com>
1704
1705         Don't trigger a Windows build when the top-level ChangeLog changes
1706
1707         r79320 is an example of a change that shouldn't have built on Windows, but did, because of
1708         this bug.
1709
1710         Fixes <http://webkit.org/b/54957> Windows bots build when top-level ChangeLog changes, but
1711         shouldn't
1712
1713         Reviewed by Anders Carlsson.
1714
1715         * Scripts/webkitpy/common/config/build.py:
1716         (_should_file_trigger_build): Modified the ChangeLog pattern to match the top-level
1717         ChangeLog, and re-sorted it in the list.
1718
1719         * Scripts/webkitpy/common/config/build_unittest.py:
1720         (ShouldBuildTest): Added a test that ChangeLogs don't trigger builds.
1721
1722 2011-02-22  Sheriff Bot  <webkit.review.bot@gmail.com>
1723
1724         Unreviewed, rolling out r78691.
1725         http://trac.webkit.org/changeset/78691
1726         https://bugs.webkit.org/show_bug.cgi?id=54956
1727
1728         Leopard Debug is successfully building again (Requested by
1729         aroben on #webkit).
1730
1731         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1732         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1733
1734 2011-02-22  Ilya Tikhonovsky  <loislo@chromium.org>
1735
1736         Reviewed by Pavel Feldman.
1737
1738         Web Inspector: flakyness of inspector tests.
1739         https://bugs.webkit.org/show_bug.cgi?id=54729
1740
1741         As far as we have the protocol with sequence numbers we can simplify test harness support
1742         and drop out chomium specific methods.
1743
1744         * DumpRenderTree/DumpRenderTree.gypi:
1745         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
1746         (DRTDevToolsAgent::sendMessageToInspectorFrontend):
1747         (DRTDevToolsAgent::runtimePropertyChanged):
1748         (DRTDevToolsAgent::asyncCall):
1749         (DRTDevToolsAgent::call):
1750         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
1751         * DumpRenderTree/chromium/DRTDevToolsCallArgs.cpp: Removed.
1752         * DumpRenderTree/chromium/DRTDevToolsCallArgs.h: Removed.
1753         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
1754         (DRTDevToolsClient::sendMessageToBackend):
1755         (DRTDevToolsClient::asyncCall):
1756         (DRTDevToolsClient::call):
1757         * DumpRenderTree/chromium/DRTDevToolsClient.h:
1758
1759 2011-02-22  Andras Becsi  <abecsi@webkit.org>
1760
1761         Reviewed by Laszlo Gombos.
1762
1763         [Qt] Redesign the build system
1764         https://bugs.webkit.org/show_bug.cgi?id=51339
1765
1766         Part 2.
1767
1768         Build WebCore as a static library, compile the WebKit API and WebKit2 API
1769         in a final step and link to WebKit2, WebCore and JSC libraries to fix
1770         linking issues resulting from stripped away symbols.
1771
1772         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Remove superfluous includepaths.
1773
1774 2011-02-21  Lukasz Slachciak  <l.slachciak@samsung.com>
1775
1776         Reviewed by Antonio Gomes.
1777
1778         [EFL] Restored spatial navigation option in EWebLauncher since #18662 bug is closed
1779         https://bugs.webkit.org/show_bug.cgi?id=54806
1780
1781         * EWebLauncher/main.c:
1782         (on_key_down): calling ewk_view_setting_spatial_navigation_set when "F12" pressed added.
1783
1784 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
1785
1786         Reviewed by Kent Tamura.
1787
1788         nrwt: revert behavior on 10.5 to "old-threads". It's possible
1789         that r79062 has introduced a bunch of flakiness, although I
1790         can't explain why right now. This change should restore the
1791         old behavior, and then we can look at the build logs to see
1792         if the tests stop timing out or being flaky.
1793
1794         https://bugs.webkit.org/show_bug.cgi?id=54925
1795
1796         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1797         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1798         * Scripts/webkitpy/layout_tests/port/mac.py:
1799         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1800
1801 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
1802
1803         Reviewed by Dimitri Glazkov.
1804
1805         nrwt: fix typo causing us to try and repeatedly shut down the servers
1806         https://bugs.webkit.org/show_bug.cgi?id=54904
1807
1808         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
1809
1810 2011-02-21  Alejandro G. Castro  <alex@igalia.com>
1811
1812         Reviewed by Csaba Osztrogonác.
1813
1814         REGRESSION(r79157):
1815         plugins/get-url-with-javascript-destroying-plugin.html make the
1816         next test crash
1817         https://bugs.webkit.org/show_bug.cgi?id=54863
1818
1819         * GNUmakefile.am:
1820
1821 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
1822
1823         Reviewed by Ojan Vafai.
1824
1825         test-webkitpy: stop skipping a bunch of tests on win32
1826         https://bugs.webkit.org/show_bug.cgi?id=54788
1827
1828         * Scripts/webkitpy/common/net/testoutputset_unittest.py:
1829         * Scripts/webkitpy/test/main.py:
1830
1831 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
1832
1833         Reviewed by Ojan Vafai.
1834
1835         Fix various bugs related to handling of dir separators and real
1836         vs. fake filesystems that were causing test-webkitpy to fail
1837         tests on win32. These bugs have a variety of causes but are
1838         lumped together because they should be easy to review.
1839
1840         https://bugs.webkit.org/show_bug.cgi?id=54700
1841
1842         * Scripts/webkitpy/common/net/testoutput.py:
1843           Here we were using os.path.sep but not escaping it prior to
1844           using in a regexp, which  caused bad things on win32.
1845         * Scripts/webkitpy/common/net/testoutput_unittest.py:
1846           Same here.
1847         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1848           Here we were using the native filesystem instead of the mock
1849           filesystem, making some tests unnecessarily platform dependent.
1850           Also we weren't setting the default configuration for the
1851           TestChromiumWinPort.
1852         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py:
1853           Here we *shouldn't* use os.path.join(), just so the code becomes
1854           less platform dependent.
1855         * Scripts/webkitpy/style/checker.py:
1856           Here we should be using os.path.sep instead of "/"; the tests
1857           require the native filesystem.
1858         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1859           Here the underlying code is using os.path.sep, so we probably
1860           should as well.
1861         * Scripts/webkitpy/style/test_expectations_unittest.py:
1862           Here we should be using the filesystem hanging of the port
1863           object (which is the mocked filesystem) rather than the native
1864           one.
1865
1866 2011-02-20  Anders Carlsson  <andersca@apple.com>
1867
1868         Reviewed by Maciej Stachowiak.
1869
1870         Crash when a plug-in requests a javascript: url that destroys the plug-in
1871         https://bugs.webkit.org/show_bug.cgi?id=54837
1872         <rdar://problem/9005475>
1873
1874         Add new plug-in test.
1875
1876         * DumpRenderTree/DumpRenderTree.gypi:
1877         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1878         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
1879         (PluginTest::NPN_GetURL):
1880         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
1881         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURLDestroyingPlugin.cpp: Added.
1882         (GetURLWithJavaScriptURLDestroyingPlugin::GetURLWithJavaScriptURLDestroyingPlugin):
1883         (GetURLWithJavaScriptURLDestroyingPlugin::NPP_New):
1884         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
1885         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1886
1887 2011-02-19  Patrick Gansterer  <paroga@webkit.org>
1888
1889         Reviewed by Andreas Kling.
1890
1891         Make WinCE a core builder
1892         https://bugs.webkit.org/show_bug.cgi?id=54801
1893
1894         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1895         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1896
1897 2011-02-19  Lukasz Slachciak  <l.slachciak@samsung.com>
1898
1899         Reviewed by Andreas Kling.
1900
1901         Fixed problem with launching EWebLauncher with "run-launcher --efl". Incorrect path referring to non-existing .libs directory ws set.
1902         https://webkit.org/b/54778
1903
1904         * Scripts/webkitdirs.pm:
1905
1906 2011-02-18  David Levin  <levin@chromium.org>
1907
1908         Reviewed by Eric Seidel.
1909
1910         check-webkit-style: Misses brace style error when the line with the { has a } in it.
1911         https://bugs.webkit.org/show_bug.cgi?id=54769
1912
1913         * Scripts/webkitpy/style/checkers/cpp.py: Fixed the check for the close brace
1914           to only look after the last open brace, so that the open brace in this line
1915           "} else {" will still be able to trigger the error.
1916         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added related unit tests.
1917
1918 2011-02-18  Zan Dobersek  <zandobersek@gmail.com>
1919
1920         Reviewed by Martin Robinson.
1921
1922         [Gtk] Re-enable meter tag support
1923         https://bugs.webkit.org/show_bug.cgi?id=54762
1924
1925         Meter tag is supported on the Gtk port, enable it by default.
1926
1927         * Scripts/build-webkit:
1928
1929 2011-02-18  David Levin  <levin@chromium.org>
1930
1931         Reviewed by Eric Seidel.
1932
1933         check-webkit-style falsely complains about WebKitGTK+ public headers
1934         https://bugs.webkit.org/show_bug.cgi?id=54650
1935
1936         * Scripts/webkitpy/style/checker.py: Add some exceptions for the gtk files.
1937         * Scripts/webkitpy/style/checkers/cpp.py: Restrict the checks for WEBKIT_API
1938           to the chromium directory (and improved the checks slightly).
1939         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added corresponding tests.
1940
1941 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
1942
1943         Reviewed by James Robinson.
1944
1945         new-run-webkit-tests: enable multiple processes by default on
1946         mac. This change removes the artificial restrictions we placed
1947         on the # of child processes we used with the old threading
1948         model, and switches to the new message-based model and multiple
1949         processes, where available. If multiple processes are not
1950         available (Leopard / Python 2.5), then we use the 'inline' model
1951         instead of the 'old-threads' model on the 'mac' port or one
1952         process on the 'chromium-mac' port. We need additional testing
1953         to see if the new 'threads' model works reliably and is worth
1954         supporting, or if we should just live with things being slightly
1955         slow.
1956
1957         https://bugs.webkit.org/show_bug.cgi?id=54596
1958
1959         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1960         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1961         * Scripts/webkitpy/layout_tests/port/mac.py:
1962         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1963         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
1964
1965 2011-02-18  Andrew Wilson  <atwilson@chromium.org>
1966
1967         Unreviewed, rolling out r79047.
1968         http://trac.webkit.org/changeset/79047
1969         https://bugs.webkit.org/show_bug.cgi?id=54596
1970
1971         Broke canary bots - please watch the canaries the next time
1972         you land this
1973
1974         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1975         * Scripts/webkitpy/layout_tests/port/mac.py:
1976
1977 2011-02-18  Emil A Eklund  <eae@chromium.org>
1978
1979         Reviewed by Darin Adler.
1980
1981         Crash in EventHandler::sendContextMenuEventForKey
1982         https://bugs.webkit.org/show_bug.cgi?id=54495
1983
1984         Add support for the context menu key (VK_APPS) to EventSender for
1985         platforms that support the key.
1986
1987         * DumpRenderTree/chromium/EventSender.cpp:
1988         (EventSender::keyDown):
1989         * DumpRenderTree/gtk/EventSender.cpp:
1990         (keyDownCallback):
1991         * DumpRenderTree/qt/EventSenderQt.cpp:
1992         (EventSender::keyDown):
1993         * DumpRenderTree/win/EventSender.cpp:
1994         (keyDownCallback):
1995
1996 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
1997
1998         Reviewed by Ojan Vafai.
1999
2000         new-run-webkit-tests: enable multiple processes by default on mac
2001         https://bugs.webkit.org/show_bug.cgi?id=54596
2002
2003         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2004         * Scripts/webkitpy/layout_tests/port/mac.py:
2005
2006 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2007
2008         Reviewed by Adam Roben.
2009
2010         Skip pretty patch unit tests if ruby isn't installed.
2011
2012         https://bugs.webkit.org/show_bug.cgi?id=54699
2013
2014         * Scripts/webkitpy/common/prettypatch_unittest.py:
2015
2016 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2017
2018         Reviewed by Adam Roben.
2019
2020         Skip mac.test_skipped_file_paths() when running on win32; the
2021         test has hardcoded '/' directory separators and since it's a
2022         port-specific test and we have coverage on other ports it's not
2023         worth it to make the test more generic.
2024
2025         https://bugs.webkit.org/show_bug.cgi?id=54698
2026
2027         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2028
2029 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2030
2031         Reviewed by Adam Roben.
2032
2033         Fix webkitpy.layout_tests.port.config_unittest to work on Win32.
2034
2035         https://bugs.webkit.org/show_bug.cgi?id=54703
2036
2037         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
2038
2039 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2040
2041         Reviewed by Adam Roben.
2042
2043         Change the default port we're mocking out on win32 when we
2044         run mock_drt_unittest. Normally we'd default to the
2045         port-specific default, but that would be "win", which doesn't
2046         currently work.
2047
2048         https://bugs.webkit.org/show_bug.cgi?id=54708
2049
2050         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2051
2052 2011-02-18  Kenneth Russell  <kbr@google.com>
2053
2054         Reviewed by James Robinson.
2055
2056         [chromium] Clean up initialization of minimum timer interval
2057         https://bugs.webkit.org/show_bug.cgi?id=54772
2058
2059         Use new GetForegroundTabTimerInterval function in webkit_support
2060         to reset the minimum timer interval before each test.
2061
2062         * DumpRenderTree/chromium/WebViewHost.cpp:
2063         (WebViewHost::reset):
2064
2065 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2066
2067         Reviewed by Adam Roben.
2068
2069         r78494 fixed the chromium_win unittest failing under win32,
2070         but not cygwin. This fixes cygwin as well.
2071
2072         https://bugs.webkit.org/show_bug.cgi?id=54652
2073
2074         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2075
2076 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
2077
2078         Reviewed by Ojan Vafai, Mihai Parparita.
2079
2080         new-run-webkit-tests: r78522 made it impossible to create a
2081         'chromium-win' or 'chromium-mac' port; they would always get the
2082         version tacked on to the end. It turns out that we actually need
2083         to be able to create these ports in order for
2084         rebaseline-chromium-webkit-tests to work correctly given the
2085         broken coverage of our layout bots. As soon as we get SL and
2086         Win 7 bots running reliably on the canaries, we can get rid of
2087         this change. Note that we only need to change the chromium_mac
2088         and win ports because linux doesn't yet use versions and that
2089         script doesn't work with non-chromium ports.
2090
2091         Also, this change fixes a bug where we were using
2092         "platform/chromium-win-win7" and
2093         "platform/chromium-mac-snowleopard" as the baseline dirs for
2094         Win 7 and SL; we should've been using the generic directories
2095         instead since they are the newest versions of the os's.
2096
2097         https://bugs.webkit.org/show_bug.cgi?id=54691
2098
2099         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2100         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2101         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2102         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2103         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2104
2105 2011-02-16  MORITA Hajime  <morrita@google.com>
2106
2107         Reviewed by Kent Tamura.
2108
2109         [Chromium][DRT] some PostDelayedTask() accesses possibly dangling this pointer.
2110         http://webkit.org/b/53899
2111
2112         Replaced static invocation wrappers with MethodTask objects,
2113         which are removed on WebViewHost destructor.
2114         
2115         * DumpRenderTree/chromium/WebViewHost.cpp:
2116         (WebViewHost::requestCheckingOfText):
2117         (WebViewHost::scheduleAnimation):
2118         (WebViewHost::closeWidgetSoon):
2119         * DumpRenderTree/chromium/WebViewHost.h:
2120         (WebViewHost::HostMethodTask::HostMethodTask):
2121         (WebViewHost::HostMethodTask::runIfValid):
2122         (WebViewHost::taskList):
2123
2124 2011-02-07  Hayato Ito  <hayato@chromium.org>
2125
2126         Reviewed by Ojan Vafai.
2127
2128         Get rid of code which writes test results from test_type's
2129         compare_output() method.
2130         Writing test results now happens in single_test_runner, using
2131         newly introduced test_result_writer module.
2132
2133         Subsequent patches will eliminate test_type/* classes.
2134
2135         https://bugs.webkit.org/show_bug.cgi?id=53908
2136
2137         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2138         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py: Added.
2139         * Scripts/webkitpy/layout_tests/test_types/image_diff.py:
2140         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
2141         * Scripts/webkitpy/layout_tests/test_types/text_diff.py:
2142
2143 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2144
2145         Reviewed by Adam Roben.
2146
2147         Add blacklist of remaining webkitpy/* modules that fail on
2148         Win32, so that test-webkitpy will at least run cleanly.
2149
2150         https://bugs.webkit.org/show_bug.cgi?id=54709
2151
2152         * Scripts/webkitpy/layout_tests/test/main.py:
2153
2154 2011-02-17  Simon Fraser  <simon.fraser@apple.com>
2155
2156         Reviewed by Sam Weinig.
2157
2158         Set a preference to ensure that scrollbars are in a predictable
2159         state for DRT.
2160
2161         * DumpRenderTree/mac/DumpRenderTree.mm:
2162         (resetDefaultsToConsistentValues):
2163
2164 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2165
2166         Reviewed by Adam Roben.
2167
2168         test-webkitpy doesn't work with the multiprocessing module on win32
2169         or cygwin, so we skip it for now.
2170
2171         https://bugs.webkit.org/show_bug.cgi?id=54520
2172
2173         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
2174         * Scripts/webkitpy/layout_tests/layout_package/run_webkit_tests_unittest.py:
2175
2176 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2177
2178         Reviewed by Adam Roben.
2179
2180         Fix version, baseline_search_path for chromium-win-win7. Also
2181         fix chromium_win_unittests, and fix port_testcase.make_port()
2182         to correctly use the options passed into it and only set
2183         results_directory() if necessary.
2184
2185         https://bugs.webkit.org/show_bug.cgi?id=54652
2186
2187         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2188         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2189         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2190
2191 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
2192
2193         Reviewed by Ojan Vafai.
2194
2195         Fix a hang in new-run-webkit-tests when run with
2196         --dry-run --worker-model=processes
2197
2198         https://bugs.webkit.org/show_bug.cgi?id=54595
2199
2200         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
2201         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2202
2203 2011-02-17  Kevin Ollivier  <kevino@theolliviers.com>
2204
2205         [wx] Build fixes after recent changes.
2206
2207         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2208         (LayoutTestController::setMinimumTimerInterval):
2209         * wx/build/settings.py:
2210
2211 2011-02-17  Csaba Osztrogonác  <ossy@webkit.org>
2212
2213         Reviewed by Andreas Kling.
2214
2215         [Qt] ImageDiff sometimes hangs
2216         https://bugs.webkit.org/show_bug.cgi?id=54641
2217
2218         * DumpRenderTree/qt/ImageDiff.cpp: Clear actualImage and baselineImage after all comparison.
2219         (main):
2220
2221 2011-02-17  Peter Varga  <pvarga@webkit.org>
2222
2223         Reviewed by Csaba Osztrogonác.
2224
2225         The run-sunspider --v8-suite doesn't work
2226         https://bugs.webkit.org/show_bug.cgi?id=54660
2227
2228         * Scripts/run-sunspider:
2229
2230 2011-02-17  Gabor Rapcsanyi  <rgabor@webkit.org>
2231
2232         Reviewed by Csaba Osztrogonác.
2233
2234         [NRWT] Fix environment setup on Qt port.
2235         https://bugs.webkit.org/show_bug.cgi?id=54549
2236
2237         * Scripts/webkitpy/layout_tests/port/qt.py:
2238
2239 2011-02-16  Yael Aharon  <yael.aharon@nokia.com>
2240
2241         Reviewed by Andreas Kling.
2242
2243         [Qt] Enable tiled backing store by default in MiniBrowser.
2244         https://bugs.webkit.org/show_bug.cgi?id=54581
2245
2246         Instead of a flag to turn it on, a new flag is defined to turn
2247         tiled backing store off.
2248
2249         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2250         (MiniBrowserApplication::handleUserOptions):
2251         * MiniBrowser/qt/MiniBrowserApplication.h:
2252         (WindowOptions::WindowOptions):
2253
2254 2011-02-16  Eric Seidel  <eric@webkit.org>
2255
2256         Reviewed by Mihai Parparita.
2257
2258         flaky test archive uploads are always 22 bytes long
2259         https://bugs.webkit.org/show_bug.cgi?id=54593
2260
2261         Speculative fix for the 22-bytes bug.
2262
2263         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
2264
2265 2011-02-16  Anna Cavender  <annacc@chromium.org>
2266
2267         Reviewed by Mihai Parparita.
2268
2269         http/tests/media should be served over HTTP (not from local file)
2270         https://bugs.webkit.org/show_bug.cgi?id=54028
2271
2272         Removed special file:// treatment of http/tests/media
2273
2274         * Scripts/old-run-webkit-tests:
2275         * Scripts/webkitpy/layout_tests/port/base.py:
2276
2277 2011-02-16  Mihai Parparita  <mihaip@chromium.org>
2278
2279         Reviewed by Ojan Vafai.
2280
2281         ChangeLog.parse_latest_entry_from_file does not handle rolled over ChangeLogs
2282         https://bugs.webkit.org/show_bug.cgi?id=54609
2283         
2284         r78737 rolled over ChangeLogs, and webkit-patch complains when landing
2285         patches such as this one that are the first to add an entry to the
2286         ChangeLog (since they don't see a second date line to know where the
2287         entry ends).
2288
2289         * Scripts/webkitpy/common/checkout/changelog.py:
2290         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
2291
2292 == Rolled over to ChangeLog-2011-02-16 ==