Google Git
Sign in
webkit / WebKit / a86b6aaaaf89e7ed066a31b4f0fde7cb9537972a / . / Tools / ChangeLog
blob: a1d3d36908e3e0117e090a48bfe1160eba600f9e [file] [log] [blame]
2011-07-01 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r90213.
http://trac.webkit.org/changeset/90213
https://bugs.webkit.org/show_bug.cgi?id=63792
The SL bot has an errant HTTP server holding port 8080
(Requested by abarth on #webkit).
* Scripts/run-webkit-tests:
2011-07-01 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix WebKit2 GTK build after r90163.
* MiniBrowser/gtk/WebBundle/WebBundleMain.c:
(didCreatePage):
2011-07-01 Adam Barth <abarth@webkit.org>
Exit faster when we get a keyboard interrupt. Previously, we'd open
Safari if you ^C NRWT after it found a failure.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-07-01 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Switch Snow Leopard over to new-run-webkit-tests (except for wk2)
https://bugs.webkit.org/show_bug.cgi?id=63786
We're down to one bug blocking the "move to NRWT" bug:
https://bugs.webkit.org/show_bug.cgi?id=34984
so we're trying transitioning Snow Leopard to NRWT as an experiment.
If we run into any blocking issues we'll of course roll this back out
(or anyone reading this should feel encouraged to).
* Scripts/run-webkit-tests:
2011-07-01 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r90192.
http://trac.webkit.org/changeset/90192
https://bugs.webkit.org/show_bug.cgi?id=63788
Appears to have caused NRWT on Chromium WebKit Vista to hang
(Requested by abarth on #webkit).
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2011-06-30 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Remove the concept of "being wedged" from new-run-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=63767
Worker processes shouldn't ever become wedged. My understanding is
that this code was originally motivated by the old threading-based
design but no longer servers any purpose.
Note: If we actually have a problem with the test harness getting
stuck, buildbot will kill us.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
2011-06-30 Adam Barth <abarth@webkit.org>
The win32 implementation of fork is sad about forking test-webkitpy.
Hopefully this will heal the bot.
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2011-06-30 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Update new-run-webkit-tests now that the wk2 shared skipped list has moved
https://bugs.webkit.org/show_bug.cgi?id=63780
The shared list was moved from mac-wk2 to wk2 as part of
http://trac.webkit.org/changeset/90125.
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-06-30 Kenichi Ishibashi <bashi@chromium.org>
Reviewed by Darin Adler.
Style Checker should flag removal of "developmentRegion = English;" from project.pbxproj.
https://bugs.webkit.org/show_bug.cgi?id=62022
Adds XcodeProjectFileChecker to check removal of "developmentRegion".
* Scripts/webkitpy/style/checker.py: Added XcodeProjectFileChecker as a checker for *.pbxproj.
* Scripts/webkitpy/style/checkers/xcodeproj.py: Added.
* Scripts/webkitpy/style/checkers/xcodeproj_unittest.py: Added.
2011-06-30 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
nrwt: make sharding tests needing locks less hard-coded
https://bugs.webkit.org/show_bug.cgi?id=63112
This change also changes the manager logic so that it will
drop the server lock as soon as all of the shards requiring
the lock have completed.
This change includes some minor namespace/import changes in the
unit tests, and also makes the Manager a new-style object, which
it should've been all along.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2011-06-30 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Remove threaded mode from new-run-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=63771
This mode is not used and is buggy. Rather than carry around a bunch
of unused buggy code, we should rip it out and focus on the
multiprocess implementation.
* Scripts/webkitpy/common/system/executive.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/server_process.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2011-06-30 Adam Barth <abarth@webkit.org>
Reviewed by Darin Adler.
Clean up output from new-run-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=63759
Printing messages from the child process looks super ugly because of
the way the pretty-printer works. Printing a blank line first is a
hack, but it makes things at least partially sane.
Also, handle the case where calling sample throws an exception.
* Scripts/webkitpy/layout_tests/port/server_process.py:
* Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
2011-06-27 Diego Gonzalez <diegohcg@webkit.org>
Reviewed by Antonio Gomes.
[Qt] DRT: Provide LayoutTestController::setDefersLoading and ::goBack support
https://bugs.webkit.org/show_bug.cgi?id=63489
Unskipped:
loader/navigation-while-deferring-loads.html
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::goBack):
(LayoutTestController::setDefersLoading):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-06-30 Adam Barth <abarth@webkit.org>
Reviewed by Dirk Pranke.
new-run-webkit-tests should stop when I ^C it
https://bugs.webkit.org/show_bug.cgi?id=63763
It still doesn't stop fast enough, but at least with this patch it
doesn't run more tests.
Also, I tweaked the UI that happens when there's a keyboard interrupt
to not be amazingly ugly. There's still a garbage "Te" left on the
console, but that's a problem for another day.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-06-30 Adam Roben <aroben@apple.com>
Make TestFailures never treat a too-many-failures build as a last-passing build for a test
Fixes <http://webkit.org/b/63758> TestFailures sometimes incorrectly says that a build was
the last passing build for a test, when in fact the test wasn't run due to too many failures
causing old-run-webkit-tests to exit early
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): If there were too many
failures in this build, assume that all the tests failed in the last build would have failed
in this build, too, had they been run. Introduced some new local variables to reduce typing,
and used them throughout.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Fix a regression from r89841 where we wouldn't
correctly remember whether old-run-webkit-tests exited early due to too many failures for
builds older than r89610. Also removed an unused parameter while I was at it.
2011-06-30 Adam Roben <aroben@apple.com>
Make TestFaiulres only load old-results directories as needed
Fixes <http://webkit.org/b/63752> Tester pages on TestFailures page load very slowly
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getBuildNames): Moved up into the API section of the class. Now just
calls through to _getBuildNamesFromResultsDirectory.
(Builder.prototype.getOldBuildNames): Added. Just calls through to
_getBuildNamesFromResultsDirectory.
(Builder.prototype._getBuildNamesFromResultsDirectory): Renamed from getBuildNames. Now
takes the directory URL as an argument and only fetches that single URL.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer.prototype.start): Moved most logic to _analyzeBuilds. First
analyzes builds from Builder.getBuildNames, then from Builder.getOldBuildNames if needed.
(LayoutTestHistoryAnalyzer.prototype._analyzeBuilds): Moved logic here from start. (Most
changes are just indentation.) Now takes a callback to call when we've finished analyzing
all builds in buildNames so that we can try to fetch more build names if needed.
2011-06-30 Mark Rowe <mrowe@apple.com>
Reviewed by Anders Carlsson.
Update MiniBrowser and WebKitTestRunner to use the client version constants.
* MiniBrowser/gtk/BrowserWindow.c:
(browserWindowLoaderClientInit):
(browserWindowUIClientInit):
* MiniBrowser/gtk/WebBundle/WebBundleMain.c:
(didCreatePage):
(WKBundleInitialize):
* MiniBrowser/gtk/main.c:
(createWKContextWithInjectedBundle):
* MiniBrowser/mac/AppDelegate.m:
(-[BrowserAppDelegate init]):
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]):
* MiniBrowser/mac/WebBundle/WebBundleMain.m:
(WKBundleInitialize):
* MiniBrowser/win/BrowserView.cpp:
(BrowserView::create):
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::initialize):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage):
(WTR::TestController::initialize):
2011-06-30 Adam Roben <aroben@apple.com>
Identify revisions that modified failing tests on TestFailures page
Fixes <http://webkit.org/b/63716> TestFailures page should try to infer which commit caused
a failure by looking at commit logs
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(.existing-bugs-list, .suspect-revisions-list): Made this rule apply to the suspect
revisions list, too.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
(Trac): Initialize our memory cache.
(Trac.prototype.getCommitDataForRevisionRange): Added. Fetches the commit log in RSS form,
then parses out the revision number, Trac's idea of the commit title, and the files modified
by the commit. (To get the modified files, we rely on the commit message including a
prepare-ChangeLog-style file list.)
(Trac.prototype.logURL): Added a new formatAsRSS parameter.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
(removePathExtension): Added. Returns a new string with the last dot and everything after it
removed.
(sorted): Added sortFunction parameter.
(Node.prototype.removeAllChildren): Added. Does what it says.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayBuilder): Updated for change to _domForRegressionRange.
(ViewController.prototype._domForRegressionRange): Changed argument order to match
_domForNewAndExistingBugs. Now takes the list of failing tests and uses it to figure out
which revisions modified the failing tests and displays the suspect revisions in a list.
2011-06-30 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Rename should_kill_dump_render_tree to driver_needs_restart in preparation for supporting WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=63743
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-06-30 Martin Robinson <mrobinson@igalia.com>
Reviewed by Anders Carlsson.
[GTK] Crash observed with nspluginwrapper and flash
https://bugs.webkit.org/show_bug.cgi?id=62249
Added a TestNetscapePlugin test which verifies that WebKit properly
handles situations where InvalidateRect is called with a null instance.
* DumpRenderTree/TestNetscapePlugIn/Tests/x11/CallInvalidateRectWithNullNPPArgument.cpp: Added.
(CallInvalidateRectWithNullNPPArgument::CallInvalidateRectWithNullNPPArgument):
(CallInvalidateRectWithNullNPPArgument::NPP_New):
* GNUmakefile.am: Add the new file to sources list.
2011-06-30 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Remove redundant tests in TestFailuresTest
https://bugs.webkit.org/show_bug.cgi?id=63742
These tests are all already covered by TestFailuresTest.test_loads.
I also added the missing "T" from TestFailuresest. :)
* Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
2011-06-30 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
REGRESSION(r84294): new-run-webkit-tests results.html generate links to diffs.html or diff.png that don't exist
https://bugs.webkit.org/show_bug.cgi?id=63494
This patch does two things:
1) Refactor diff_image to return the image diff instead of writing the
diff to a file. This fixes a bunch of hacks and disentangles a bunch
of code.
2) When there's a checksum mismatch but not image diff, we no longer
report an IMAGE failure to results.html. That fixes the bug in
question because results.html won't try to link to a non-existent image
diff.
* Scripts/webkitpy/common/system/executive_mock.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
* Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2011-06-28 Hans Wennborg <hans@chromium.org>
Reviewed by Tony Chang.
IndexedDB: Prepare for running layout tests with LevelDB
https://bugs.webkit.org/show_bug.cgi?id=63593
The TestShell must always provide a temporary folder for LevelDB,
not just when the backing store type is overridden.
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::setOverrideIndexedDBBackingStore):
(LayoutTestController::clearAllDatabases):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
* DumpRenderTree/chromium/TestShell.h:
2011-06-30 Zsolt Fehér <h490267@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[WK2] Create a general wk2 specific LayoutTests/platform/wk2 platform
qt-wk2, win-wk2, mac-wk2 use this common Skipped list too
https://bugs.webkit.org/show_bug.cgi?id=61785
* Scripts/old-run-webkit-tests:
2011-06-30 Adam Roben <aroben@apple.com>
Use a colon to separate test names from failure info on TestFailures
This matches how the "Possibly Flaky Tests" list works, and is much prettier when the
failure info itself includes parentheses.
Fixes <http://webkit.org/b/63722> Crashing symbols on TestFailures page look ugly due to
double parentheses
Reviewed by Darin Adler.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForFailedTest): Use a colon instead of surrounding the failure
info in parentheses.
2011-06-30 Adam Roben <aroben@apple.com>
Robustify PersistentCache against cached data that contains the string ': '
Fixes <http://webkit.org/b/63721> Exceptions thrown while loading TestFailures page due to
passing a bad string to JSON.parse
Reviewed by Darin Adler.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js:
(PersistentCache._parseDateAndJSONFromString): Manually split the string on the separator
using indexOf and substring. (String.prototype.split always splits on all instances of the
separator.)
2011-06-30 Adam Roben <aroben@apple.com>
Include the crashing symbol in crash logs links on TestFailures
Fixes <http://webkit.org/b/63465> Links to crash logs on TestFailures page should include
the crashing symbol (like links in results.html do)
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Bumped the cache version because we now store
crashing symbols for crashing tests. Renamed testsForResultTable to parseResultTable because
it now returns more than just the test names. Specifically, it now looks for crash log links
and extracts the crashing symbol name from them. Updated callers of parseResultTable to
match its new behavior. Changed to store the crashing symbol along with the failure type in
the data we pass to the callback for tests which crashed.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(code): Make <code> elements a little smaller because their contents can be quite long.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForFailureDiagnosis): Include the crashing symbol inside a
<code> element in the link, if there is a crashing symbol.
2011-06-30 Adam Roben <aroben@apple.com>
Use objects instead of strings to represent a test result in TestFailures code
This will eventually allow us to store more than just the type of failure for each test.
(E.g., we can store the name of the crashing symbol for tests which crashed.)
Prep work for <http://webkit.org/b/63465> Links to crash logs on TestFailures page should
include the crashing symbol (like links in results.html do)
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.failureDiagnosisTextAndURL): Changed to expect a testResult object
instead of just a failureType string.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
(FlakyLayoutTestDetector.prototype.incorporateTestResults): Changed to store a
testResult-like object for passing tests.
(FlakyLayoutTestDetector.prototype.flakinessExamples): Changed to expect testResult-like
objects.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Store a version number along with the cached data
so we can throw away cached data that's in an old format. Store a testResult object for each
test instead of just its failure type.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForFailedTest):
(ViewController.prototype._domForFailureDiagnosis):
Changed to expect testResult objects instead of failureType strings.
2011-06-30 Adam Roben <aroben@apple.com>
Show full commit logs when visiting Trac from TestFailures page
Fixes <http://webkit.org/b/63715> Links to Trac on TestFailures page
should show full commit logs
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
(Trac.prototype.logURL): Added new 'showFullCommitLogs' parameter. When specified, we add
verbose=on to the log URL to tell Trac to show full commit logs.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForRegressionRange): Specify true for showFullCommitLogs.
2011-06-30 Adam Roben <aroben@apple.com>
Add a link to file a bug about TestFailures to the TestFailures page
Fixes <http://webkit.org/b/63466> TestFailures page should include a link to file a bug
about TestFailures itself
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(aside): Added some styles for the new link. It is positioned in the upper-right and
hopefully not to glaringly obvious.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype.loaded): Set up our main content element and auxiliary UI.
(ViewController.prototype._displayBuilder):
(ViewController.prototype._displayTesters):
Modify the main content element instead of the entire body, as there are other things in the
body that we want to preserve.
(ViewController.prototype._domForAuxiliaryUIElements): Create and return an <aside> element
that contains a link to file a bug about the TestFailures page.
2011-06-30 Adam Roben <aroben@apple.com>
Ignore callbacks that correspond to old navigations of the TestFailures page
Fixes <http://webkit.org/b/63711> TestFailures page sometimes ignores back/forward
navigations
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer.prototype.start): Check the return value of the callback. If the
callback returns false, stop fetching more data.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController): Initialize new _navigationID member. This member is used to assign each
navigation by the user a unique ID.
(ViewController.prototype.parseHash): Increment the navigation ID.
(ViewController.prototype._displayBuilder):
(ViewController.prototype._displayTesters):
At the start of these functions, record the current navigation ID. Bail out from callbacks
if the navigation ID has changed, indicating that the user has navigated elsewhere.
2011-06-30 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Add back/forward menu to MiniBrowser toolbar
https://bugs.webkit.org/show_bug.cgi?id=63445
* MiniBrowser/gtk/BrowserWindow.c:
(browser_window_init): Use a GtkMenuToolButton for navigation
widgets.
(browserWindowConstructed): Get the page BackForwardList.
(browserWindowHistoryItemActivated): Go to activated history item.
(browserWindowHistoryItemSelected): Show the url of currently
selected history item in the statusbar.
(createGtkActionFromBackForwardItem): Create a GtkAction from a
WKBackForwardListItemRef.
(browserWindowCreateMenuItemFromBackForwardItem): Create a
GtkMenuItem from a WKBackForwardListItemRef.
(browserWindowCreateBackForwardMenu): Create a GtkMenu with the
given history items.
(browserWindowUpdateNavigationActions): Update navigation widgets
sensitivity and history menus.
(didChangeBackForwardList): Call
browserWindowUpdateNavigationActions().
(browserWindowLoaderClientInit): Add didChangeBackForwardList
callback.
2011-06-30 Dmitry Lomov <dslomov@google.com>
Reviewed by Adam Roben.
https://bugs.webkit.org/show_bug.cgi?id=63660
This enables unit-tests step on Windows and MacOS SnowLeopard.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-30 Balazs Kelemen <kbalazs@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt][WK2] Implement the rest of WTR::PlatformWebView
https://bugs.webkit.org/show_bug.cgi?id=63630
* WebKitTestRunner/qt/PlatformWebViewQt.cpp:
(WTR::PlatformWebView::windowFrame):
(WTR::PlatformWebView::setWindowFrame):
2011-06-29 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
test-webkitpy is spamming lots of messages about updating test expectations
https://bugs.webkit.org/show_bug.cgi?id=63680
This patch changes the passing_run() function to capture the
output to stdout and stderr and assert that it is empty, and
also specifies --no-new-test-results by default to avoid getting
"generating baseline" messages for the expected-missing results.
Also check that we are logging messages when we generate new
baselines.
Also fix a minor bug in the way test_tolerance() was
checking for whether the tests were passing or not.
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2011-06-29 Yuta Kitamura <yutak@chromium.org>
Reviewed by Kent Tamura.
DRT: Make Hixie76WebSocketProtocolEnabled preference flag configurable from LayoutTestController
https://bugs.webkit.org/show_bug.cgi?id=63532
LayoutTestController::overridePreference() will be used to change the value of
the Hixie76WebSocketProtocolEnabled preference flag from layout tests. In this way,
we can avoid writing a lot of boilerplate code in each port's DumpRenderTree, but
at least we need to reset preference values in DumpRenderTree every time we start
a new test because preference values changed in some test should not be leaked
to the next test.
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::overridePreference):
The name of preference key is defined in WebKit/{mac/WebView,win}/WebPreferenceKeysPrivate.h.
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::reset):
Set the default value. This will be called from TestShell::resetTestController().
(WebPreferences::applyTo):
* DumpRenderTree/chromium/WebPreferences.h:
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetDefaultsToConsistentValues): Set the default value.
* DumpRenderTree/win/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues): Ditto.
2011-06-29 Kenichi Ishibashi <bashi@chromium.org>
Reviewed by Adam Barth.
check-webkit-style fails to detect https://webkit.org/b/XXXXX as a bug number
https://webkit.org/b/57360
Modify the regular expression to match https scheme.
* Scripts/webkitpy/common/checkout/changelog_unittest.py: Added a test case for https://webkit.org/b/XXXXX.
* Scripts/webkitpy/common/config/urls.py: Modified bug_url_short to match https scheme.
2011-06-29 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
new-run-webkit-tests: images page switch actual and expected images before they're loaded
https://bugs.webkit.org/show_bug.cgi?id=63199
We now preload the images and keep the DOM nodes in memory instead of
going back to the MemoryCache all the time. (Also, remove some nutty
</img> HTML.)
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2011-06-29 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
new-run-webkit-tests should tell you when it creates new expectations
https://bugs.webkit.org/show_bug.cgi?id=63675
Unforunately, this patch creates a bunch of test-webkitpy spam because
we don't understand how Python logging works well enough to stop the
spam. :(
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2011-06-29 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Refactor WebKitDriver.run_test into smaller pieces
https://bugs.webkit.org/show_bug.cgi?id=63673
There is no functional change here. Just moving code around.
This is in preparation for adding support for #CRASHED and #CRASHED - WebProcess
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-06-29 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Simplify skipped file finding in preparation for adding wk2 skipped list fallback
https://bugs.webkit.org/show_bug.cgi?id=63501
Re-landing after fixing a bug in WebKitPort.baseline_search_path which broke
the rebaselineserver unit tests. Also fixed nits brought up by Dirk after previous landing.
The goal was to add support for wk2 skipped lists.
However, I found that our skipped list computation was a
manual hack (which only worked for the mac port).
So I fixed a FIXME to move the skipped list finding
into WebKitPort instead of MacPort.
Doing so required the concept of a "port_name", since previously
the only name accessible from a port object was name()
which includes many things beyond "mac" or "chromium", etc.
Eventually I believe we'll want to re-think the way that we pass
in a port_name argument to Port subclasses and expect them to parse
it. But for now I just added a cls.port_name variable which contains
the static information needed to compute wk2 names as well as
compute Skipped list fallback which works for Mac/Win/Qt and Gtk.
In order to test my new _skipped_file_search_paths method, I
fixed another FIXME by making it return relative paths.
I also fixed the test_expectations_path code in WebKitPort to use port_name.
It was using name() which would return PORT-VERSION so MacPort was overriding
it to use just PORT. After fixing test_expectations_path to use port_name
(and making it aware of webkit2) I was able to remove the MacPort implementation.
* Scripts/webkitpy/layout_tests/port/base.py:
- Add port_name() to access "mac", since name() returns "mac-leopard" etc.
- Document that real_name() seems to have no purpose.
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
- Add _parse_port_name(), eventually we might call this from WebKitPort directly.
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
- Add _parse_port_name.
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
- Add _parse_port_name.
* Scripts/webkitpy/layout_tests/port/gtk.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
- Move Skipped-file finding code down to WebKitPort
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/qt.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-06-29 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r90070.
http://trac.webkit.org/changeset/90070
https://bugs.webkit.org/show_bug.cgi?id=63676
Broke test-webkitpy rebaseline tests (Requested by eseidel2 on
#webkit).
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/gtk.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/qt.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
* Scripts/webkitpy/layout_tests/port/win.py:
* Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
2011-06-29 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
new-run-webkit-tests doesn't support sample-on-timeout
https://bugs.webkit.org/show_bug.cgi?id=56731
* Scripts/webkitpy/layout_tests/port/server_process.py:
* Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-06-29 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Simplify skipped file finding in preparation for adding wk2 skipped list fallback
https://bugs.webkit.org/show_bug.cgi?id=63501
The goal was to add support for wk2 skipped lists.
However, I found that our skipped list computation was a
manual hack (which only worked for the mac port).
So I fixed a FIXME to move the skipped list finding
into WebKitPort instead of MacPort.
Doing so required the concept of a "port_name", since previously
the only name accessible from a port object was name()
which includes many things beyond "mac" or "chromium", etc.
Eventually I believe we'll want to re-think the way that we pass
in a port_name argument to Port subclasses and expect them to parse
it. But for now I just added a cls.port_name variable which contains
the static information needed to compute wk2 names as well as
compute Skipped list fallback which works for Mac/Win/Qt and Gtk.
In order to test my new _skipped_file_search_paths method, I
fixed another FIXME by making it return relative paths.
I also fixed the test_expectations_path code in WebKitPort to use port_name.
It was using name() which would return PORT-VERSION so MacPort was overriding
it to use just PORT. After fixing test_expectations_path to use port_name
(and making it aware of webkit2) I was able to remove the MacPort implementation.
* Scripts/webkitpy/layout_tests/port/base.py:
- Add port_name() to access "mac", since name() returns "mac-leopard" etc.
- Document that real_name() seems to have no purpose.
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
- Add _parse_port_name(), eventually we might call this from WebKitPort directly.
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
- Add _parse_port_name.
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
- Add _parse_port_name.
* Scripts/webkitpy/layout_tests/port/gtk.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
- Move Skipped-file finding code down to WebKitPort
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/qt.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-06-29 Adam Barth <abarth@webkit.org>
Reviewed by Dirk Pranke.
new-run-webkit-tests complains about missing pixel results instead of plopping down new expectations
https://bugs.webkit.org/show_bug.cgi?id=38063
This patch changes new-run-webkit-tests to match old-run-webkit-tests
in generating new expected results when the expected results are
missing.
There are still a couple details that are different:
1) Image baselines aren't generated unless you run with -p.
2) Render tree dumps are places in the cross-platform directory instead
of the platform-specific directory.
I'm inclined to deal with both of these issues in follow-up patches.
* Scripts/webkitpy/common/net/layouttestresults.py:
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_results.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2011-06-29 Adam Roben <aroben@apple.com>
Teach TestFailures to detect possibly flaky tests and list them separately
Fixes <http://webkit.org/b/61061> <rdar://problem/9452796> TestFailures page blames
arbitrary revisions for breaking flaky tests
Reviewed by Dan Bates.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js: Added.
(FlakyLayoutTestDetector): This class identifies flaky tests when given the test results
from various builds (in reverse-chronological order).
(FlakyLayoutTestDetector.prototype.incorporateTestResults): Detects flaky tests. Tests move
monotonically through three states: LastSeenFailing, LastSeenPassing, and PossiblyFlaky.
(FlakyLayoutTestDetector.prototype.flakinessExamples): Finds examples of flakiness for the
given test. Essentially, finds all the transitions from passing to failing (or vice-versa)
and puts them in an array in reverse-chronological order.
(FlakyLayoutTestDetector.prototype.get possiblyFlakyTests): Returns all tests we've detected
to be possibly flaky.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer): Initialize new members.
(LayoutTestHistoryAnalyzer.prototype.start): Now passes the callback an object with two
properties: history and possiblyFlaky. history holds the data this function used to pass to
the callback, while possiblyFlaky lists all tests that might be flaky and examples of their
flakiness. Updated documentation comment to match.
(LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Now uses a
FlakyLayoutTestDetector to identify possibly flaky tests. Any possibly flaky tests are
removed from the failure history, since when they started failing is no longer meaningful.
We tell our caller to keep calling until all current failures have been explained and we've
gone through 5 builds without any new flaky tests being identified.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
(sorted): New helper function to return a sorted copy of an array.
(Array.prototype.findLast): New helper function. Like findFirst, but finds the last item
that matches the predicate.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayBuilder): Updated for change in the object passed to us by
the analyzer. Now puts the list of possibly flaky tests after the failure history.
(ViewController.prototype._domForFailedTest): Moved some code from here...
(ViewController.prototype._domForFailureDiagnosis): ...to here.
(ViewController.prototype._domForPossiblyFlakyTests): New function, builds up a list of
possibly flaky tests and examples of their flakiness and returns it.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
FlakyLayoutTestDetector.js.
2011-06-29 Eric Seidel <eric@webkit.org>
Adam says cowboys don't review (or unit test).
* Scripts/webkitpy/tool/commands/download.py:
2011-06-29 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Make port/config.py mockable for easier unit testing
https://bugs.webkit.org/show_bug.cgi?id=63661
Fix an exception from the unit tests.
* Scripts/webkitpy/layout_tests/port/config.py:
2011-06-29 Brent Fulgham <bfulgham@webkit.org>
Unreviewed build correction.
* DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops:
Link against zlib1.dll, not the static zlib.lib.
2011-06-29 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Make port/config.py mockable for easier unit testing
https://bugs.webkit.org/show_bug.cgi?id=63661
Right now config.py leaks real system paths for
webkit_base_dir due to using __file__. This patch
adds a filsystem.py wrapper around __file__ so that
we don't have to jump through hoops in webkit_unittest.py
to avoid leaking local paths to the unit test results.
There are probably many more places where we should now
use filesystem.path_for_module, but starting with just this one.
* Scripts/webkitpy/common/system/filesystem.py:
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/layout_tests/port/config.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
* Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
2011-06-29 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Remove duplicate methods in filesystem.py
https://bugs.webkit.org/show_bug.cgi?id=63658
Looks like there was a bad merge at some point.
I also removed a bunch of redundant docstrings.
* Scripts/webkitpy/common/system/filesystem.py:
2011-06-29 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Pass --no-new-test-results to NRWT on build.webkit.org
https://bugs.webkit.org/show_bug.cgi?id=63653
This is in preparation for https://bugs.webkit.org/show_bug.cgi?id=38063.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-29 Xan Lopez <xlopez@igalia.com>
Reviewed by Adam Barth.
[GTK] overlapping drag&drop tests fail on NRWT
https://bugs.webkit.org/show_bug.cgi?id=57640
Make a GTK test driver that spawns one Xvfb instance per
thread. This avoids bad interactions in DnD tests between threads.
* Scripts/webkitpy/layout_tests/port/gtk.py:
2011-06-29 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r89888.
http://trac.webkit.org/changeset/89888
https://bugs.webkit.org/show_bug.cgi?id=63609
"Seems to be causing merging issues on the dashboard, e.g.
http://test-
results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20chromium.org&tests=fast%2Fcanvas"
(Requested by aboxhall on #webkit).
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2011-06-29 Philippe Normand <pnormand@igalia.com>
Reviewed by Martin Robinson.
[webkitpy] Executive.running_pids() sometimes doesn't correctly process "ps" output
https://bugs.webkit.org/show_bug.cgi?id=63624
* Scripts/webkitpy/common/system/executive.py: In some cases the
ps output lines can contain one or more leading white-spaces, so
strip it before splitting the pid and command name.
2011-06-29 Noel Gordon <noel.gordon@gmail.com>
Reviewed by Kent Tamura.
[Chromium] [DRT] ImageDiff should compile in VS2010
https://bugs.webkit.org/show_bug.cgi?id=63614
* DumpRenderTree/chromium/ImageDiff.cpp: Include <iterator> for back_inserter.
2011-06-27 Diego Gonzalez <diegohcg@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt] DRT support for setInteractiveFormValidationEnabled
https://bugs.webkit.org/show_bug.cgi?id=63496
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::DumpRenderTree):
2011-06-28 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Make new-run-webkit-tests -2 actually use WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=63572
new-run-webkit-tests -2 works with this change. However a zillion
tests fail since it can't find the right skipped lists for wk2.
The Skipped-list fixing is covered in bug 63501.
* Scripts/webkitpy/layout_tests/port/gtk.py:
* Scripts/webkitpy/layout_tests/port/qt.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-06-28 Ryosuke Niwa <rniwa@webkit.org>
Reviewed by Adam Roben.
Add more machines to Chromium Mac Release and Chromium Mac Release (Tests)
https://bugs.webkit.org/show_bug.cgi?id=63500
Added google-mac-3 and google-mac-4.
* BuildSlaveSupport/build.webkit.org-config/config.json:
2011-06-28 Eric Seidel <eric@webkit.org>
Reviewed by Dirk Pranke.
Remove webkitpy's tiger support now that WebKit no longer supports tiger
https://bugs.webkit.org/show_bug.cgi?id=63568
I don't think our support actually fully worked anyway. Regardless
tiger is dead now (thankfully)!
I also unwrapped some short lines and added a couple FIXMEs while
touring the code.
No new tests, only removing old ones which tested our Tiger support.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/builders.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/tool/commands/rebaseline.py:
* Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
2011-06-28 Eric Seidel <eric@webkit.org>
Unreviewed, just fixing a typo.
Fix misspellings of "thead" to "thread".
* Scripts/webkitpy/common/checkout/scm/git.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-06-28 Eric Seidel <eric@webkit.org>
Reviewed by Tony Chang.
new-run-webkit-tests needs a --webkit-test-runner option
https://bugs.webkit.org/show_bug.cgi?id=63439
This was accidentally fixed in bug 63501 as well.
The root problem here was that the webkit unit tests were not using
enough mocks. I've fixed them here to use more mocks
which should unbreak windows.
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-06-28 Eric Seidel <eric@webkit.org>
Reviewed by Tony Chang.
new-run-webkit-tests should not run mthml tests when MHTML support is disabled
https://bugs.webkit.org/show_bug.cgi?id=63549
This way mhtml tests won't show up as "missing" results on all non-chromium platforms.
I happen to go looking for what code was deciding that .mht was a supported
extension, found test_files.py and then removed some redundant doc-strings and
if branching.
* Scripts/webkitpy/layout_tests/port/test_files.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-06-28 Roland Steiner <rolandsteiner@chromium.org>
Reviewed by Tony Chang.
Fix filesystem_unittest.FileSystemTest test_read_and_write_file()
https://bugs.webkit.org/show_bug.cgi?id=63514
- change write_text_file to write_binary_file
- check that text_path and binary_path are actually existant files
- change cleanup block from 'except' to 'finally'
* Scripts/webkitpy/common/system/filesystem_unittest.py:
2011-06-28 Adam Roben <aroben@apple.com>
Roll out r89924 and r89925
I didn't mean to commit them!
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
NRWT should wait for ReportCrash
https://bugs.webkit.org/show_bug.cgi?id=63495
When a test crashes on Mac, ReportCrash confuses us into thinking the
test times out.
I'm not sure how to test this in a meaningful way in our current test
harness...
* Scripts/webkitpy/common/system/crashlogs.py:
* Scripts/webkitpy/common/system/crashlogs_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/server_process.py:
* Scripts/webkitpy/tool/commands/queries.py:
2011-06-27 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r89881.
http://trac.webkit.org/changeset/89881
https://bugs.webkit.org/show_bug.cgi?id=63513
Makes Qt massively slow (Requested by abarth on #webkit).
* Scripts/webkitpy/common/system/crashlogs.py:
* Scripts/webkitpy/common/system/crashlogs_unittest.py:
* Scripts/webkitpy/common/system/executive.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/server_process.py:
* Scripts/webkitpy/tool/commands/queries.py:
2011-06-27 Alice Boxhall <aboxhall@chromium.org>
Reviewed by Ojan Vafai.
Convert json_results_generator.py to output version 4 JSON.
https://bugs.webkit.org/show_bug.cgi?id=60869
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
NRWT should wait for ReportCrash
https://bugs.webkit.org/show_bug.cgi?id=63495
When a test crashes on Mac, ReportCrash confuses us into thinking the
test times out.
I'm not sure how to test this in a meaningful way in our current test
harness...
* Scripts/webkitpy/common/system/crashlogs.py:
* Scripts/webkitpy/common/system/crashlogs_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/server_process.py:
* Scripts/webkitpy/tool/commands/queries.py:
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
new-run-webkit-tests should upload crash logs
https://bugs.webkit.org/show_bug.cgi?id=55907
NRWT used to have a different name for crash logs. I've changed the
name to patch ORWT.
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2011-06-27 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
new-run-webkit-tests needs a --webkit-test-runner option
https://bugs.webkit.org/show_bug.cgi?id=63439
NRWT doesn't actually know how to run with the WebKitTestRunner yet
but it does have a --webkit-test-runner option and will build WebKitTestRunner correctly.
There is a bunch of other little cleanup in this patch which I added as I
took a tour through all of our hard-coded DumpRenderTree strings.
* Scripts/webkitpy/common/net/layouttestresults.py: Added FIXME.
* Scripts/webkitpy/common/system/outputcapture.py: Fixed spacing to pass PEP8.
* Scripts/webkitpy/layout_tests/layout_package/manager.py: Made the FIXME slightly stronger (bad bug).
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: test_shell is dead.
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Unwrapped silly wrapping.
* Scripts/webkitpy/layout_tests/port/base.py: Removed two dead methods.
* Scripts/webkitpy/layout_tests/port/config.py:
- build_dumprendertree had no business in this class, removed it.
- Exposed _FLAGS_FROM_CONFIGURATIONS through flag_for_configuration()
- Exposed _script_path as script_path() (this probably belongs elsewhere).
* Scripts/webkitpy/layout_tests/port/config_unittest.py:
- Moved these tests to webkit_unittests.
* Scripts/webkitpy/layout_tests/port/webkit.py:
- Now respects the --webkit-test-runner option.
- setup_test_run is empty in base.py too, no need to override it.
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
- Test the new hotness.
- Bad, bad, bad! The old code was using a real Executive during unit-testing!
I think this code is still hitting disk during the unit tests. :(
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
- Add the option.
* Scripts/webkitpy/tool/bot/queueengine_unittest.py:
- Remove optional args option.
* Scripts/webkitpy/tool/commands/commandtest.py:
- Remove optional args option.
* Scripts/webkitpy/tool/commands/queues_unittest.py:
- Remove optional args option.
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Dirk Pranke.
new-run-webkit-tests does not report stderr output
https://bugs.webkit.org/show_bug.cgi?id=37739
Almost all the infrastructure appears to be in place to report stderr
in results.html. We just need to wire up a bool to tell results.html
to create the hyperlink.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/layout_package/test_results.py:
* Scripts/webkitpy/layout_tests/port/base.py:
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Dirk Pranke.
webkitpy unit tests should have more descriptive names than just "Test"
https://bugs.webkit.org/show_bug.cgi?id=63479
Luckily we kept these classes in separate namespaces in the harness so
we were actually running them all even though they had the same name.
* Scripts/webkitpy/common/system/stack_utils_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
2011-06-27 Greg Simon <gregsimon@chromium.org>
No review necessary.
Add email addr to contributor list
https://bugs.webkit.org/show_bug.cgi?id=63478
* Scripts/webkitpy/common/config/committers.py:
2011-06-27 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r89834.
http://trac.webkit.org/changeset/89834
https://bugs.webkit.org/show_bug.cgi?id=63476
Broke test-webkitpy (Requested by abarth on #webkit).
* Scripts/webkitpy/common/net/layouttestresults.py:
* Scripts/webkitpy/common/system/outputcapture.py:
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/config.py:
* Scripts/webkitpy/layout_tests/port/config_unittest.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/tool/bot/queueengine_unittest.py:
* Scripts/webkitpy/tool/commands/commandtest.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
2011-06-27 Jeff Miller <jeffm@apple.com>
Reviewed by Adam Roben.
Stop copying obsolete WebKit libraries on Windows
https://bugs.webkit.org/show_bug.cgi?id=63474
Don't copy obsolete WebKit library directories and files on Windows, they no longer exist and just generate extraneous warnings. These include the CharacterSets directory, dnssd.dll, and icudt40.dll.
* DumpRenderTree/win/DumpRenderTreePostBuild.cmd:
* MiniBrowser/MiniBrowserPostBuild.cmd:
* WebKitTestRunner/win/WebKitTestRunnerPostBuild.cmd:
2011-06-27 Ryosuke Niwa <rniwa@webkit.org>
Add Wyatt Carss to the list of non-committer contributors.
* Scripts/webkitpy/common/config/committers.py:
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
webkitpy should understand crash logs
https://bugs.webkit.org/show_bug.cgi?id=63468
We're planning to use this functionality to upload crash logs along
with test results for new-run-webkit-tests.
* Scripts/webkitpy/common/system/crashlog.py: Added.
* Scripts/webkitpy/common/system/crashlog_unittest.py: Added.
* Scripts/webkitpy/common/system/executive.py:
* Scripts/webkitpy/common/system/executive_unittest.py:
* Scripts/webkitpy/common/system/filesystem.py:
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/tool/commands/queries.py:
2011-06-27 Adam Roben <aroben@apple.com>
Make LayoutTestResultsLoader cache whether old-run-webkit-tests exited early due to too many
failures
Fixes <http://webkit.org/b/63470> TestFailures page for a particular builder forgets
old-run-webkit-tests exited early after reload
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Store both the set of failing tests and whether
old-run-webkit-tests exited early due to too many failures in PersistentCache.
2011-06-27 Adam Roben <aroben@apple.com>
Make TestFailures load build names from build.webkit.org/old-results too
Build results are periodically moved from results to old-results. This change makes those
builds still visible to TestFailures.
Fixes <http://webkit.org/b/63453> TestFailures page doesn't show information for builds that
have been moved to build.webkit.org/old-results
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getBuildNames): Extracted code to parse build names from a
build.webkit.org directory listing into a separate function. Instead of omitting .zip files,
we now only include directory entries whose names are parseable as build names (since
old-results sometimes contains other random files/directories from who knows what). We now
fetch both results and old-results (with a FIXME about loading old-results on demand),
extract build names from each, and concatenate the two sets of names.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js:
(WebKitBuildbot.prototype.parseBuildName): Changed to return null when the build name isn't
parseable, rather than throwing an exception.
2011-06-26 Adam Roben <aroben@apple.com>
Store analyzed history directly in LayoutTestHistoryAnalyzer instead of in a generic cache
It seems silly to have a whole generic cache just to store a single object.
Fixes <http://webkit.org/b/63407> LayoutTestHistoryAnalyzer has a generic cache but doesn't
need one
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer): Replaced this._cache with this._history.
(LayoutTestHistoryAnalyzer.prototype.start): Ditto.
(LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Ditto, and removed the
no-longer-needed history argument.
2011-06-26 Adam Roben <aroben@apple.com>
Move some inappropriate code out of Builder
Fixes <http://webkit.org/b/63406> TestFailure page's Builder class has a bunch of code that
isn't really about a builder
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getBuildNames): Renamed from _getBuildNames, since it now needs to be
called by other objects.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
Added. Code came from Builder.
(LayoutTestHistoryAnalyzer):
(LayoutTestHistoryAnalyzer.prototype.start):
(LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory):
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
Added. Code came from Builder.
(LayoutTestResultsLoader):
(LayoutTestResultsLoader.prototype.start):
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayBuilder): Changed to use LayoutTestHistoryAnalyzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in new
scripts.
2011-05-17 Nat Duca <nduca@chromium.org>
Reviewed by James Robinson.
[chromium] Implement CCLayerTreeHost and CCLayerTreeHostImpl portions of threaded compositor
https://bugs.webkit.org/show_bug.cgi?id=58408
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::paintInvalidatedRegion):
2011-06-27 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
new-run-webkit-tests needs a --webkit-test-runner option
https://bugs.webkit.org/show_bug.cgi?id=63439
NRWT doesn't actually know how to run with the WebKitTestRunner yet
but it does have a --webkit-test-runner option and will build WebKitTestRunner correctly.
There is a bunch of other little cleanup in this patch which I added as I
took a tour through all of our hard-coded DumpRenderTree strings.
* Scripts/webkitpy/common/net/layouttestresults.py: Added FIXME.
* Scripts/webkitpy/common/system/outputcapture.py: Fixed spacing to pass PEP8.
* Scripts/webkitpy/layout_tests/layout_package/manager.py: Made the FIXME slightly stronger (bad bug).
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: test_shell is dead.
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Unwrapped silly wrapping.
* Scripts/webkitpy/layout_tests/port/base.py: Removed two dead methods.
* Scripts/webkitpy/layout_tests/port/config.py:
- build_dumprendertree had no business in this class, removed it.
- Exposed _FLAGS_FROM_CONFIGURATIONS through flag_for_configuration()
- Exposed _script_path as script_path() (this probably belongs elsewhere).
* Scripts/webkitpy/layout_tests/port/config_unittest.py:
- Moved these tests to webkit_unittests.
* Scripts/webkitpy/layout_tests/port/webkit.py:
- Now respects the --webkit-test-runner option.
- setup_test_run is empty in base.py too, no need to override it.
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
- Test the new hotness.
- Bad, bad, bad! The old code was using a real Executive during unit-testing!
I think this code is still hitting disk during the unit tests. :(
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
- Add the option.
* Scripts/webkitpy/tool/bot/queueengine_unittest.py:
- Remove optional args option.
* Scripts/webkitpy/tool/commands/commandtest.py:
- Remove optional args option.
* Scripts/webkitpy/tool/commands/queues_unittest.py:
- Remove optional args option.
2011-06-27 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
[WK2] WebKitTestRunner should reset FrameFlattening flag to disable before each test
https://bugs.webkit.org/show_bug.cgi?id=63457
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
2011-06-27 Raphael Kubo da Costa <kubo@profusion.mobi>
Reviewed by Kent Tamura.
[EFL] DRT: Add an ewk_view specialization for EFL's DumpRenderTree.
It will be plugged into the build system as soon as the rest of the
DRT code is fully upstreamed.
https://bugs.webkit.org/show_bug.cgi?id=63086
* DumpRenderTree/efl/DumpRenderTreeView.cpp: Added.
(onConsoleMessage):
(onJavaScriptAlert):
(onJavaScriptConfirm):
(onJavaScriptPrompt):
(drtViewTiledAdd):
* DumpRenderTree/efl/DumpRenderTreeView.h: Added.
2011-06-27 Leandro Pereira <leandro@profusion.mobi>
Reviewed by Kent Tamura.
[EFL] Add an EventSender implementation to EFL's DumpRenderTree along
with support files.
Build system changes will be made as soon as the whole DumpRenderTree
is upstreamed.
https://bugs.webkit.org/show_bug.cgi?id=61962
* DumpRenderTree/efl/EventSender.cpp: Added.
(setEvasModifiers):
(translateMouseButtonNumber):
(sendMouseEvent):
(sendClick):
(scheduleAsynchronousClickCallback):
(updateClickCount):
(modifierFromJSValue):
(modifiersFromJSValue):
(mouseDownCallback):
(mouseUpCallback):
(mouseMoveToCallback):
(evasMouseEventFromHorizontalAndVerticalOffsets):
(mouseScrollByCallback):
(continuousMouseScrollByCallback):
(keyPadNameFromJSValue):
(keyNameFromJSValue):
(keyDownCallback):
(scalePageByCallback):
(getClass):
(makeEventSender):
* DumpRenderTree/efl/EventSender.h: Added.
2011-06-27 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Remove evil uses of hasattr
https://bugs.webkit.org/show_bug.cgi?id=63430
For some reason these classes believe that they may be called with
various flavors of "option" elements and so carefully check to make
sure that the options element has their option before checking it.
We had a set_option_default method which was never called, so I made it
do what callsites seemed to want it to do and replaced 3 callers
who previously used hasattr manually to use set_option_default instead.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-06-27 Kent Tamura <tkent@chromium.org>
Reviewed by Hajime Morita.
[Chromium] WebViewHost should be destructed before LayoutTestController, etc.
https://bugs.webkit.org/show_bug.cgi?id=63427
~WebViewHost() navigates to about:blank, so it calls some
callbacks and LayoutTestController functions.
* DumpRenderTree/chromium/TestShell.h:
Move m_printer upper and move m_webViewHost lower.
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Move server config files to webkitpy.layout_tests.servers
https://bugs.webkit.org/show_bug.cgi?id=63436
The config files belong with the servers.
* Scripts/webkitpy/layout_tests/port/httpd2.pem: Removed.
* Scripts/webkitpy/layout_tests/port/lighttpd.conf: Removed.
* Scripts/webkitpy/layout_tests/servers/http_server.py:
* Scripts/webkitpy/layout_tests/servers/httpd2.pem: Copied from Tools/Scripts/webkitpy/layout_tests/port/httpd2.pem.
* Scripts/webkitpy/layout_tests/servers/lighttpd.conf: Copied from Tools/Scripts/webkitpy/layout_tests/port/lighttpd.conf.
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Move websocket_server and http_server to webkitpy.layout_package.servers
https://bugs.webkit.org/show_bug.cgi?id=63434
These classes do not represent ports. They represent servers.
* Scripts/new-run-webkit-httpd:
* Scripts/new-run-webkit-websocketserver:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/http_server.py: Removed.
* Scripts/webkitpy/layout_tests/port/http_server_base.py: Removed.
* Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py: Removed.
* Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
* Scripts/webkitpy/layout_tests/servers/http_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server.py.
* Scripts/webkitpy/layout_tests/servers/http_server_base.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server_base.py.
* Scripts/webkitpy/layout_tests/servers/http_server_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py.
* Scripts/webkitpy/layout_tests/servers/websocket_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/websocket_server.py.
2011-06-27 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Remove Chromium and ChromiumMac implementations of driver_name now that everything uses DumpRenderTree
https://bugs.webkit.org/show_bug.cgi?id=63428
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Move apache_http_server.py to webkitpy.layout_package.servers
https://bugs.webkit.org/show_bug.cgi?id=63433
The apache_http_server.py is a server, not a port.
* Scripts/webkitpy/layout_tests/port/apache_http_server.py: Removed.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/servers: Added.
* Scripts/webkitpy/layout_tests/servers/__init__.py: Added.
* Scripts/webkitpy/layout_tests/servers/apache_http_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/apache_http_server.py.
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Move rebaseline-chromium-webkit-tests.py to webkitpy.to_be_moved
https://bugs.webkit.org/show_bug.cgi?id=63432
This code isn't in the dependency cone of new-run-webkit-tests. It's
unclear what it's final home should be, but this patch moves it out of
layout_tests to simply the package.
* Scripts/rebaseline-chromium-webkit-tests:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Removed.
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Removed.
* Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py.
* Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py.
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Move update_webgl_conformance_tests.py to webkitpy.to_be_moved
https://bugs.webkit.org/show_bug.cgi?id=63429
This package as nothing to do with the layout_tests package.
* Scripts/update-webgl-conformance-tests:
* Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py: Removed.
* Scripts/webkitpy/layout_tests/update_webgl_conformance_tests_unittest.py: Removed.
* Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py.
* Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/update_webgl_conformance_tests_unittest.py.
2011-06-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Move deduplicate_tests.py into webkitpy.to_be_moved
https://bugs.webkit.org/show_bug.cgi?id=63426
deduplicate_tests.py doesn't haven anything to do with layout tests.
It's really a separate script. This patch moves it to the (new)
to_be_moved package so that it doesn't distract us when working on NRWT.
* Scripts/deduplicate-tests:
* Scripts/webkitpy/layout_tests/deduplicate_tests.py: Removed.
* Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py: Removed.
* Scripts/webkitpy/test/main.py:
* Scripts/webkitpy/to_be_moved: Added.
* Scripts/webkitpy/to_be_moved/__init__.py: Added.
* Scripts/webkitpy/to_be_moved/deduplicate_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/deduplicate_tests.py.
* Scripts/webkitpy/to_be_moved/deduplicate_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py.
2011-06-26 Adam Barth <abarth@webkit.org>
Rubber-stamped by Eric Seidel.
Remove empty directory.
* Scripts/webkitpy/layout_tests/test_types: Removed.
2011-06-25 Naoki Takano <honten@chromium.org>
Reviewed by Eric Seidel.
Change my entry from contributors to committers.
https://bugs.webkit.org/show_bug.cgi?id=63043
* Scripts/webkitpy/common/config/committers.py: Change my e-mail from takano.naoki@gmail.com to honten@chromium.org. And move the entry to commiters from contributors.
2011-06-25 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
[Chromium] Update WebKit for WEBKIT_FRAME_TO_DOCUMENT_API_MOVE
https://bugs.webkit.org/show_bug.cgi?id=63394
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::queueLoad):
2011-06-25 Brent Fulgham <bfulgham@webkit.org>
Unreviewed build correction after r89426.
* WebKitTestRunner/Configurations/WebKitTestRunnerCFLite.vsprops:
Link with Cairo DLL for new functionality.
* WebKitTestRunner/cairo/TestInvocationCairo.cpp: Include compatibility
code for snprintf.
2011-06-25 Dimitri Glazkov <dglazkov@chromium.org>
Reviewed by Adam Barth.
run-bindings-tests does not work on Chromium Windows.
https://bugs.webkit.org/show_bug.cgi?id=63382
* BuildSlaveSupport/build.webkit.org-config/master.cfg: Removed run-bindings-tests as a build step for cr-win.
2011-06-25 Dan Bernstein <mitz@apple.com>
Try to fix the Qt and Chromium builds.
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::setEncodedAudioData):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2011-06-24 Dan Bernstein <mitz@apple.com>
Reviewed by Anders Carlsson.
Add an option to enable legacy rounding hacks
https://bugs.webkit.org/show_bug.cgi?id=63363
Added layoutTestController.allowRoundingHacks() and implemented it
on OS X.
* DumpRenderTree/LayoutTestController.cpp:
(allowRoundingHacksCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::allowRoundingHacks):
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::allowRoundingHacks):
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::allowRoundingHacks):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::allowRoundingHacks):
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::allowRoundingHacks):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::allowRoundingHacks):
2011-06-24 Adam Roben <aroben@apple.com>
Include the directory containing the failing tests in titles of bugs filed from TestFailures
when all the tests won't fit
Fixes <http://webkit.org/b/63350> Short bug titles from TestFailures page give no indication
which tests are failing
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
(longestCommonPathPrefix): Added. Given a set of paths, returns the longest common prefix
that ends in a path separator.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForNewAndExistingBugs): When we can't fit all the test names
in the title, first try to include the longest common prefix of the test names, then, if
that's still too long or there's no common prefix, fall back to not including any
information about which tests are failing.
2011-06-24 Adam Roben <aroben@apple.com>
Include links to Trac in bugs filed from TestFailures
Fixes <http://webkit.org/b/63348> Bugs filed from TestFailures page should include links to
Trac for suspect revisions
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
(Trac.prototype.changesetURL): New function, returns the URL for a particular revision in
Trac.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForNewAndExistingBugs): If there's only one suspect revision,
include a link to Trac for it in the first sentence of the description. Otherwise include a
link to the suspect range after the list of failing tests.
2011-06-24 Nat Duca <nduca@chromium.org>
Unreviewed, rolling out r89694.
http://trac.webkit.org/changeset/89694
https://bugs.webkit.org/show_bug.cgi?id=58408
Test shell still not ready for animate changes.
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::paintInvalidatedRegion):
2011-06-24 Adam Roben <aroben@apple.com>
Fix typos introduced in r89696
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getNumberOfFailingTests): Pass the result data's constituent parts to the
callback, since that's what it expects.
2011-06-24 Adam Roben <aroben@apple.com>
Make TestFailures correctly remember whether old-run-webkit-tests exited early
Fixes <http://webkit.org/b/63342> TestFailures page incorrectly claims test run ran to
completion after reload
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getNumberOfFailingTests): Instead of just storing the number of failures
in the PeristentCache, store an object that contains both the number of failures and whether
old-run-webkit-tests exited early.
2011-06-24 Adam Roben <aroben@apple.com>
Add links to regression ranges in Trac to the TestFailures page
Fixes <http://webkit.org/b/61060> <rdar://problem/9452153> TestFailures page should provide
links to regression ranges in Trac
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js: Added.
(Trac): This new class represents a particular instance of Trac for a single project.
(Trac.prototype.logURL): Returns the URL that can be used to see the log of the given
repository path between the two specified revisions
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController): Added new trac argument.
(ViewController.prototype._displayBuilder): Moved code to create the DOM for the
passed/failed builds from here...
(ViewController.prototype._domForRegressionRange): ...to here. Now also includes a link to
Trac if there are multiple suspect revisions.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
Trac.js and pass a Trac instance for trac.webkit.org to the ViewController.
2011-05-17 Nat Duca <nduca@chromium.org>
Reviewed by James Robinson.
[chromium] Implement CCLayerTreeHost and CCLayerTreeHostImpl portions of threaded compositor
https://bugs.webkit.org/show_bug.cgi?id=58408
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::paintInvalidatedRegion):
2011-06-24 Dominic Cooney <dominicc@chromium.org>
Reviewed by Dimitri Glazkov.
Convert shadow DOM-related tests to use window.internals
https://bugs.webkit.org/show_bug.cgi?id=61671
Remove ensureShadowRoot, shadowRoot, shadowPseudoId and
removeShadowRoot from layoutTestController; these have been
replaced by equivalents in window.internals instead.
* DumpRenderTree/LayoutTestController.cpp:
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-06-24 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Dirk Schulze.
Make run-bindings-tests remove its temporary directories.
https://bugs.webkit.org/show_bug.cgi?id=63320
* Scripts/run-bindings-tests:
2011-06-23 MORITA Hajime <morrita@google.com>
Reviewed by Kent Tamura.
[Chromium] Invalid write inside WebKit::FrameLoaderClientImpl::dispatchDidClearWindowObjectInWorld
https://bugs.webkit.org/show_bug.cgi?id=63216
Deleted m_webViewHost which looks leaked.
* DumpRenderTree/chromium/TestShell.h:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::~TestShell):
2011-06-24 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix WebKit2 GTK build.
* WebKitTestRunner/GNUmakefile.am:
2011-06-23 Ryuan Choi <ryuan.choi@samsung.com>
Unreviewed, add myself as committer.
* Scripts/webkitpy/common/config/committers.py:
2011-06-23 Nat Duca <nduca@chromium.org>
Unreviewed, add myself as committer.
* Scripts/webkitpy/common/config/committers.py:
2011-06-23 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: move http locking code into manager
https://bugs.webkit.org/show_bug.cgi?id=63103
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-06-23 Alok Priyadarshi <alokp@chromium.org>
Reviewed by James Robinson.
[chromium] Add support for running layout tests with accelerated rendering
https://bugs.webkit.org/show_bug.cgi?id=63274
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-06-23 Adam Roben <aroben@apple.com>
Avoid fetching JSON data when possible on TestFailures page to determine if
old-run-webkit-tests exited early
As of r89610, results.html includes information about whether old-run-webkit-tests exited
early, so we don't need to look at JSON data to get this information for builds newer than
that.
Fixes <http://webkit.org/b/63281> TestFailures page fetches build JSON for every build,
which is slow and limits history to the last build master restart
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype._getFailingTests): Moved code dealing with results.html into a new
function. If we're dealing with a build that has early-exit information in results.html,
don't bother fetching JSON data and go straight to parsing results.html. Otherwise first use
the JSON data to determine if old-run-webkit-tests exited early, and then parse results.html
(as before).
2011-06-23 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
sheriffbot can no longer restart
https://bugs.webkit.org/show_bug.cgi?id=63221
Fixed and unit tested.
* Scripts/webkitpy/tool/bot/queueengine.py:
* Scripts/webkitpy/tool/bot/sheriffircbot.py:
* Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2011-06-23 Martin Robinson <mrobinson@igalia.com>
Reviewed by Xan Lopez.
[GTK] [WebKit2] WebKitTestRunner does not properly load TestNetscapePlugin
https://bugs.webkit.org/show_bug.cgi?id=63287
Send the path to the TestNetscapePlugin to the WKTR via an environment variable
so that the plugin loads properly.
* Scripts/old-run-webkit-tests: Pass the appropriate environment variable to WKTR.
* WebKitTestRunner/gtk/TestControllerGtk.cpp:
(WTR::getEnvironmentVariableAsUTF8String): Added this helper which abstracts away some
of the work for getting an environment variable and converting it to UTF-8.
(WTR::TestController::initializeInjectedBundlePath): Use the helper.
(WTR::TestController::initializeTestPluginDirectory): Get the path from an environment variable now.
2011-06-23 Adam Roben <aroben@apple.com>
Record whether old-run-webkit-tests aborted early in results.html
Fixes <http://webkit.org/b/63280> results.html gives no indication of whether
old-run-webkit-tests stopped running tests early
Reviewed by David Kilzer.
* Scripts/old-run-webkit-tests:
(top level): Switched to heredoc syntax for printing the start of results.html for clarity.
Added a <style> element with some styles for the new message. If we stopped running early,
put the reason why at the top of results.html.
(stopRunningTestsEarlyIfNeeded): Store the message we print in the new
$stoppedRunningEarlyMessage global.
2011-06-23 Adam Roben <aroben@apple.com>
Cache some of the TestFailures page's data in localStorage
This makes reloading TestFailures much faster. Right now we only store the number of failing
tests and the list of failing tests for each build. We may choose to store more later, but
it's easy to run up against quota limits.
Fixes <http://webkit.org/b/61520> TestFailures page should take advantage of LocalStorage
APIs (or similar) to improve loading performance
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getNumberOfFailingTests): Changed to use the new PersistentCache object.
(Builder.prototype.startFetchingBuildHistory): Changed to pass whether or not we're still
fetching data to the callback.
(Builder.prototype._getFailingTests): Changed to use the new PersistentCache object. We now
store the tests in the cache just before calling the callback. (The previous code in this
function relied on being able to modify the tests object after storing it in the cache and
having the cached version be updated. This worked while it was a non-serialized cache, but
PersistentCache uses serialization.)
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js: Added.
(PersistentCache): This object wraps localStorage. It uses JSON to serialize/deserialize
values, and stores the date that each value was initially stored along with it. This is
later used for pruning the cache.
(PersistentCache.contains): Checks whether the key exists in localStorage.
(PersistentCache.get): Fetch the string we stored in localStorage and extract the original
value out of it.
(PersistentCache.set): Serialize the value, add the date to it, and store it in
localStorage. If this fails due to quota limits, empty the whole cache and try again.
(PersistentCache.prune): Delete any cached data that is deemed old enough.
(PersistentCache._addDateToJSONString): Prepend the current date to the string.
(PersistentCache._emptyCache): Delete everything from localStorage.
(PersistentCache._parseDateAndJSONFromString): Split apart the date and the JSON string and
return them.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayBuilder): Updated for change to callback signature. When
we finish fetching data, prune the PersistentCache. While I was here I also fixed a bug
where we'd never show the new bug link for tests for which we couldn't determine a passing
revision.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
PersistentCache.js.
2011-06-23 Adam Roben <aroben@apple.com>
Make finding existing bugs and filing new bugs work on TestFailures even when lots of tests are failing
One bug this fixes is <http://webkit.org/b/61660> New bug links on TestFailures page often
contain titles that are so long they are rejected by Bugzilla
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
(Bugzilla.prototype.quickSearch): Use fetchResource to POST the search query rather than
using a query string on the URL. If the search query is very long, using a query string can
cause the request to be rejected due to the URL being too long. POSTing avoids this issue.
(Bugzilla.maximumBugTitleLength): Added this constant based on Bugzilla's implementation.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(.new-bug-form): Hide the form that we secretly use to file a new bug.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
(fetchResource): Added. Code came from getResource. If we're using a GET request, add the
query parameters to the URL. Otherwise, send them as the body of the request along with the
appropriate headers.
(getResource): Now just calls through to fetchResource.
(urlEncodedQueryParameters): Added. Moved some code here...
(addQueryParametersToURL): ...from here.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForNewAndExistingBugs): Shorten the title to just mention the
number of failing tests if mentioning all the tests would make the title too long. Improved
the description for large numbers of failures by listing the tests one-per-line instead of
just having them be comma-separated, which was hard to read. Use a form to file the new bug
instead of a link so that we can POST the form data. (Using a URL with a query string can
result in the URL being too long.) The new bug link now just submits the form.
2011-06-23 Adam Roben <aroben@apple.com>
Show closed bugs on the TestFailures page in addition to open ones
Fixes <http://webkit.org/b/63194> TestFailures page should show closed bugs, too
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
(Bugzilla.prototype.quickSearch): Added code to extract the bug's status and store it in the
returned data.
(Bugzilla.isOpenStatus): New function, returns true if the passed-in status indicates that
the associated bug is still open.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
(addQueryParametersToURL): Add a missing semicolon.
(Node.prototype.appendChildren): New function, like appendChild but takes an array-like
object and appends each of the values stored within.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForNewAndExistingBugs): Prepend 'ALL' to the query so closed
bugs will be included in the results. Split the returned bugs into two sets: those which are
open, and those which are closed. Put the open bugs at the top level of the list, and the
closed bugs in a second level.
2011-06-23 Adam Roben <aroben@apple.com>
Don't count new tests as failures on the TestFailures page
Fixes <http://webkit.org/b/63254> TestFailures page calls new tests "failures", even though
they aren't
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getNumberOfFailingTests): Don't add the number of new tests to the number
of failures.
2011-06-23 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
nrwt: make TestInput objects printable
https://bugs.webkit.org/show_bug.cgi?id=63225
This makes debugging slighlty easier.
* Scripts/webkitpy/layout_tests/layout_package/test_input.py:
2011-06-22 Daniel Bates <dbates@webkit.org>
Reviewed by Adam Barth.
test-webkitpy --all errors out because scm_unittest.py can't find module checkout
https://bugs.webkit.org/show_bug.cgi?id=62943
Remove "from .checkout import Checkout" from scm_unittest.py since it's included
by Scripts/webkitpy/common/checkout/__init__.py.
* Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2011-06-22 Dominic Cooney <dominicc@chromium.org>
Reviewed by Mark Rowe.
Add window.internals to WebKit2's WebKitTestRunner.
https://bugs.webkit.org/show_bug.cgi?id=61073
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pxbproj:
* WebKitTestRunner/Configurations/Base.xcconfig:
* WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
* WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
* WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::didClearWindowForFrame):
* WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2011-06-22 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Make sheriff-bot rollout messages a little nicer
https://bugs.webkit.org/show_bug.cgi?id=63107
itertools.chain.from_iterable is new in 2.6,
use itertools.chain(*list) for 2.5 compat.
* Scripts/webkitpy/tool/bot/irc_command.py:
2011-06-22 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Make sheriff-bot robust against exceptions from commands
https://bugs.webkit.org/show_bug.cgi?id=63211
sheriff-bot was acting strange this afternoon.
We don't know if this will fix the cause, but
at least it adds some unit tests and catches
one possible cause.
* Scripts/webkitpy/tool/bot/sheriffircbot.py:
* Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2011-06-22 Mark Rowe <mrowe@apple.com>
Reviewed by David Levin.
<http://webkit.org/b/63212> TestWebKitAPI Xcode project has bogus settings since r86287
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Fix the reference to libgtest.a to
be relative to the built products directory rather than using a hard-coded path. Remove the
explicit settings of FRAMEWORK_SEARCH_PATHS, HEADER_SEARCH_PATHS and LIBRARY_SEARCH_PATHS
as they're all unnecessary. If they were necessary they should be set in the .xcconfig file
rather than in the .xcodeproj directly.
2011-06-22 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: remove obsolete port.shut_down_http_server method
https://bugs.webkit.org/show_bug.cgi?id=59993
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/gtk.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/port/qt.py:
* Scripts/webkitpy/layout_tests/port/win.py:
2011-06-22 Luke Macpherson <macpherson@chromium.org>
Reviewed by Andreas Kling.
Move macpherson from contributor list to committer list.
https://bugs.webkit.org/show_bug.cgi?id=63179
* Scripts/webkitpy/common/config/committers.py:
Move macpherson from contributor list to committer list.
2011-06-22 David Levin <levin@chromium.org>
Reviewed by Adam Barth.
check-webkit-style should detect returning (Own|Ref)Ptr instead of the Pass*Ptr version.
https://bugs.webkit.org/show_bug.cgi?id=63204
* Scripts/webkitpy/style/checkers/cpp.py: Added a check for the return value and combined
with similar code for the parameter checking.
* Scripts/webkitpy/style/checkers/cpp_unittest.py: Removed pass_ptr checks from
those done for single lines since they don't make sense in that case (variable decls look like function decls).
Removed some redundant comments (one of which was slightly wrong).
Added checks for the new functionality and minor other test changes.
2011-06-22 Nate Chapin <japhet@chromium.org>
Reviewed by Adam Barth.
Add libsoup to libWebCoreInternals build.
https://bugs.webkit.org/show_bug.cgi?id=62066
* GNUmakefile.am:
2011-06-22 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: handle missing httpd cleanly
https://bugs.webkit.org/show_bug.cgi?id=62027
We had reverted the change in r89414, so this adds it back in
(modifying check_sys_deps() in port/base.py to check for an
installed web server).
This change then fixes the change in r89414 to stub out the
check_sys_deps() in the test port so that the unit tests run correctly.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/test.py:
2011-06-22 David Levin <levin@chromium.org>
Reviewed by Adam Barth.
check-webkit-style should check for invalid uses of RefPtr/OwnPtr as parameters.
https://bugs.webkit.org/show_bug.cgi?id=63188
* Scripts/webkitpy/style/checkers/cpp.py: Added the check.
* Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests.
2011-06-22 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
Add a buildbot for testing ENABLE(CSS3_FLEXBOX)
https://bugs.webkit.org/show_bug.cgi?id=62891
* BuildSlaveSupport/build.webkit.org-config/config.json:
2011-06-22 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
nrwt: don't look for http lock when running the test port
https://bugs.webkit.org/show_bug.cgi?id=63158
* Scripts/webkitpy/layout_tests/port/mock_drt.py:
* Scripts/webkitpy/layout_tests/port/test.py:
2011-06-22 Balazs Kelemen <kbalazs@webkit.org>
Reviewed by Andreas Kling.
[Qt][WK2] Set up plugin tests
https://bugs.webkit.org/show_bug.cgi?id=63066
* WebKitTestRunner/qt/TestControllerQt.cpp:
(WTR::TestController::initializeTestPluginDirectory):
Use the QTWEBKIT_PLUGIN_PATH environment variable as the path
of the test plugin like DRT does.
2011-06-22 Eric Seidel <eric@webkit.org>
Reviewed by Ojan Vafai.
Make sheriff-bot rollout messages a little nicer
https://bugs.webkit.org/show_bug.cgi?id=63107
It annoyed me this afternoon that I had to convert sheriff-bots "r12345" revisions
into urls myself. So I have now fixed its "preparing" message to include a url.
I also figured that I should make the messages mention all of the responsible parties
so that rollouts are never surprises. If you're in the channel and were involved
in a patch, you will see if someone is using sheriff-bot to rollout a patch.
As part of doing this I also changed (and tested) _parse_args to fail-fast
when given invalid args.
* Scripts/webkitpy/tool/bot/irc_command.py:
* Scripts/webkitpy/tool/bot/irc_command_unittest.py:
* Scripts/webkitpy/tool/bot/sheriff.py:
* Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2011-06-22 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
run-webkit-websocketserver fails to stop websocket server
https://bugs.webkit.org/show_bug.cgi?id=63123
There's a ten-character change in new-run-webkit-websocketserver
that fixes the actual bug, and then we update the code in
new-run-webkit-httpd to do the actual stop call, and add more tests
(refactoring the existing test code so that it is shared).
* Scripts/new-run-webkit-httpd:
* Scripts/new-run-webkit-websocketserver:
* Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py:
2011-06-22 Adam Roben <aroben@apple.com>
Add links to existing bugs related to failing tests on TestFailures page
I changed the layout of the page a little to make it easier to read with all the new
information. Passing/failing revisions have been moved down below the list of tests to be
closer to the existing bugs and the new bug link. And each set of tests and its relevant
information is in a light gray box.
Fixes <http://webkit.org/b/61665> TestFailures page should link to existing bugs when
possible
Reviewed by Darin Adler.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js: Added.
(Bugzilla): This new class represents a single Bugzilla instance.
(Bugzilla.prototype.quickSearch): Searches Bugzilla using its Quick Search functionality,
passing the resulting bug titles and URLs to the callback when complete. If called multiple
times with the same query before the query returns, caches the callbacks so that only one
query is sent over the wire. When the query completes, all pending callbacks are called.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(#failure-history): Reduce the margin/padding on the top-level list a bit.
(#failure-history > li): Put each set of tests in a gray box, and indent most information
inside the box.
(.test-list): Unindent the list of failing tests so it is visually at the top level.
(.new-and-existing-bugs): Reduce the space at the bottom of this area so that the bottom of
each box isn't a big empty space.
(.existing-bugs-list): Use a smaller text size for existing bugs, since their titles can be
quite long.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
(addQueryParametersToURL): New function extracted from
ViewController.prototype._domForNewAndExistingBugs.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController): Take and store an optional Bugzilla instance.
(ViewController.prototype._displayBuilder): Give the top-level list an id for styling
purposes and move the list of failing tests above all other information. Only show bug
information once we've determined the most-recent passing revision for a set of tests. It's
not that useful to file a new bug before this information has been determined, and searching
for existing bugs before we've figured out which tests started failing at the same time
would end up giving you information about a bunch of unrelated tests.
(ViewController.prototype._domForNewAndExistingBugs): Renamed from _domForNewBugLink. Now
returns a DocumentFragment instead of an HTMLParagraphElement. If we don't have a Bugzilla
instance, just returns an empty DocumentFragment. Starts a search for bugs related to the
failing tests, and adds links to the bugs when the search completes.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pass a
Bugzilla instance for bugs.webkit.org to the ViewController.
2011-06-22 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Apply current view settings to newly created views in GtkLauncher
https://bugs.webkit.org/show_bug.cgi?id=63142
* GtkLauncher/main.c:
(createWebViewCb):
2011-06-22 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Use GOption to parse main arguments in GtkLauncher
https://bugs.webkit.org/show_bug.cgi?id=63060
Instead of getting the settings from the view, create a new
settings object not associated with any view so that we can parse
the command line arguments before creating the view. It will also
allows to use the same settings for more than one view if we
eventually support passing more than one uri from the command
line.
* GtkLauncher/main.c:
(parseOptionEntryCallback): Make this function static. Also use
g_ascii_strtoll/g_ascii_strtod to parse int/float values instead
of atoi/atof to provide better error messages. And always fill the
error struct when the function fails.
(getOptionEntriesFromWebKitWebSettings): Use an array of
GOptionEntry structs instead of a GArray, since that is what
g_option_group_add_entries() expects and the array size doesn't
change.
(addWebSettingsGroupToContext): Helper function to add the
websettings group to the option context.
(main): Create the option context to parse command line arguments.
2011-06-22 Martin Robinson <mrobinson@igalia.com>
Reviewed by Adam Roben.
[GTK] Implement pixel dump support for WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=58242
* DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Added.
* GNUmakefile.am: Separate out the ImageDiff bits of the GNUmakefile.
* Scripts/old-run-webkit-tests: Update to add support for the GTK+ port.
* WebKitTestRunner/GNUmakefile.am: Added Cairo specific files to the build.
* WebKitTestRunner/TestInvocation.cpp: Moved this code to the platform-independent
file so that we can share it across platforms.
(WTR::TestInvocation::compareActualHashToExpectedAndDumpResults):
* WebKitTestRunner/TestInvocation.h: Updated method list.
* WebKitTestRunner/cairo/TestInvocationCairo.cpp: Added Cairo implementation.
* WebKitTestRunner/cg/TestInvocationCG.cpp:
(WTR::TestInvocation::dumpPixelsAndCompareWithExpected): Updated to use new helper.
2011-06-22 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r89414.
http://trac.webkit.org/changeset/89414
https://bugs.webkit.org/show_bug.cgi?id=63130
It broke webkitpy tests (Requested by Ossy on #webkit).
* Scripts/webkitpy/layout_tests/port/base.py:
2011-06-22 Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu>
Reviewed by Dirk Pranke.
[NRWT] Print error message when there is no httpd present on the system
https://bugs.webkit.org/show_bug.cgi?id=62027
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
2011-06-21 Dirk Pranke <dpranke@chromium.org>
Unreviewed, build fix.
Fix crashes in new-run-webkit-tests resulting from the
change to the http_server logic in r89400. Python 2.5
on Mac 10.5 has some weird error paths.
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/common/system/executive.py:
2011-06-21 MORITA Hajime <morrita@google.com>
Unreviewed, rolling out r89401 and r89403.
http://trac.webkit.org/changeset/89401
http://trac.webkit.org/changeset/89403
https://bugs.webkit.org/show_bug.cgi?id=62970
Breaks mac build and mistakenly enables the spellcheck API
* Scripts/build-webkit:
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2011-06-21 Kent Tamura <tkent@chromium.org>
[Mac] Sort Xcode project files.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2011-06-20 MORITA Hajime <morrita@google.com>
Reviewed by Kent Tamura.
Spellcheck API should be build-able.
https://bugs.webkit.org/show_bug.cgi?id=62970
* Scripts/build-webkit:
2011-06-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: attempt #3 at fixing server startup/shutdown
https://bugs.webkit.org/show_bug.cgi?id=62829
Attempt yet again to make starting and stopping servers
reliable. It turns out that apache has races between when the
ctl process exists and when it writes/removes the pid file.
This change accounts for the races.
* Scripts/new-run-webkit-httpd:
* Scripts/webkitpy/common/system/executive.py:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-21 Dmitry Lomov <dslomov@google.com>
Reviewed by Adam Roben.
https://bugs.webkit.org/show_bug.cgi?id=63080
Unit-tests step on test-only bot is broken.
Removing the step until run-api-tests is fixed.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-20 Dmitry Lomov <dslomov@google.com>
Reviewed by David Levin.
https://bugs.webkit.org/show_bug.cgi?id=63017
"Unit tests" step on bots should provide more information in case of failure.
* BuildSlaveSupport/build.webkit.org-config/master.cfg: added --verbose flag to UnitTests step.
2011-06-20 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: order shards by name, not number of tests
https://bugs.webkit.org/show_bug.cgi?id=62753
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-06-20 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
we should only log the test names to the tests_run.txt files during layout_tests
https://bugs.webkit.org/show_bug.cgi?id=62751
* Scripts/old-run-webkit-tests:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-06-20 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
commit-queue/ews-bots should run run-bindings-tests for every patch
https://bugs.webkit.org/show_bug.cgi?id=63010
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/tool/commands/download_unittest.py:
* Scripts/webkitpy/tool/steps/runtests.py:
* Scripts/webkitpy/tool/steps/runtests_unittest.py:
* Scripts/webkitpy/tool/steps/steps_unittest.py:
2011-06-20 Chang Shu <cshu@webkit.org>
Reviewed by Adam Barth.
fast/js/sputnik/Conformance/10_Execution_Contexts/10.2_Entering_An_Execution_Context/10.2.2_Eval_Code tests fail when showModalDialog isn't supported
https://bugs.webkit.org/show_bug.cgi?id=53676
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::initialize):
2011-06-20 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
buildbots should run run-bindings-tests
https://bugs.webkit.org/show_bug.cgi?id=63001
run-bindings-tests takes 2.4 seconds on my machine, so
I expect this will not be noticeable on the bots.
I also removed tiger-specific checks since Tiger is no
longer a supported platform for webkit.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-20 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix WebKit2 GTK build after r89249.
* GtkLauncher/main.c:
(main):
2011-06-20 Pavel Feldman <pfeldman@chromium.org>
Reviewed by Yury Semikhatsky.
Web Inspector: remove LayoutTestController::setTimelineProfilingEnabled.
https://bugs.webkit.org/show_bug.cgi?id=62994
* DumpRenderTree/LayoutTestController.cpp:
(closeWebInspectorCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
* DumpRenderTree/chromium/DRTDevToolsAgent.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-06-20 Lukasz Slachciak <l.slachciak@samsung.com>
Reviewed by Martin Robinson.
[GTK] General mechanism for adjusting WebKitWebSettings in GtkLauncher.
https://bugs.webkit.org/show_bug.cgi?id=55308
* GtkLauncher/main.c: Added general mechanism for adjusting WebKitWebSettings.
(parseOptionEntryCallback): Callback for parsing option entry.
(getOptionEntriesFromWebKitWebSettings): Basing on the WebKitWebSettings properties prepare
list of option entries.
(transformStringToBoolean): Transform function for boolean convertion used by parseOptionEntryCallback.
(transformStringToInt): Transform function for int convertion used by parseOptionEntryCallback.
(transformStringToFloat): Transform function for float convertion used by parseOptionEntryCallback.
(parseAdditionalOptions): Help function parsing additional commandline options.
(main): Added call to parseAdditionalOptions.
2011-06-20 Viatcheslav Ostapenko <ostapenko.viatcheslav@nokia.com>
Reviewed by Andreas Kling.
[Qt] [WK2] Compile error in ActivateFontsQt.cpp on Mac
https://bugs.webkit.org/show_bug.cgi?id=62869
Add QFont include to fix "incomlete type QFont" error.
* WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
2011-06-19 Adam Bergkvist <adam.bergkvist@ericsson.com>
Reviewed by Martin Robinson.
[GTK] Enable Media Stream feature and make it default on
https://bugs.webkit.org/show_bug.cgi?id=60394
* Scripts/build-webkit:
2011-06-19 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r89198.
http://trac.webkit.org/changeset/89198
https://bugs.webkit.org/show_bug.cgi?id=62953
Chromium breakage on "Linux Tests" (Requested by senorblanco
on #webkit).
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2011-06-18 Daniel Bates <dbates@webkit.org>
Reviewed by Eric Seidel.
Git.push_local_commits_to_server() should use passed username and
password when we don't have cached credentials
https://bugs.webkit.org/show_bug.cgi?id=62941
Fixes a issue where Git.push_local_commits_to_server() would always look
for cached credentials regardless of whether a username and password
were passed as arguments. Instead, if given both a username and password
then we shouldn't look for cached credentials (since we were explicitly given
credentials).
* Scripts/webkitpy/common/checkout/scm/git.py:
* Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Added unit tests.
2011-06-18 Daniel Bates <dbates@rim.com>
Rubber-stamped by Eric Seidel.
Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
https://bugs.webkit.org/show_bug.cgi?id=62767
Part 3 of 3
Remove the command-line option --skip-build now that the master BuildBot has been restarted
following <http://trac.webkit.org/changeset/89138>.
Instead, specify --no-build to get analogous functionality.
* Scripts/run-javascriptcore-tests:
2011-06-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: fix test sorting
https://bugs.webkit.org/show_bug.cgi?id=58691
This changes the sorting to use a "natural" sort order such that
foo2.html sorts before foo13.html, and sorts the tests so
that all the tests in a directory are run before any test in a
subdirectory.
This change also adds a split() method to the filesystem since
that is needed to cluster files into a directory otherwise.
* Scripts/webkitpy/common/system/filesystem.py:
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2011-06-18 Evan Martin <evan@chromium.org>
Reviewed by Darin Adler.
Compiler warning in double->int conversion in TestNetscapePlugIn
https://bugs.webkit.org/show_bug.cgi?id=51554
This code is providing a JS API for removing a property by numeric ID.
We get it from JS as a double, and must convert it to an int before using it.
Test: LayoutTests/plugins/npruntime/remove-property.html
* DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
(NPRuntimeRemoveProperty::TestObject::invoke):
2011-06-18 Patrick Gansterer <paroga@webkit.org>
Reviewed by Eric Seidel.
Don't exit webkit-patch when running on Win32 python
https://bugs.webkit.org/show_bug.cgi?id=55811
Most of the basic actions work in the meantime,
so let people play with in on Win32 now.
* Scripts/webkit-patch:
2011-06-18 Alice Boxhall <aboxhall@chromium.org>
Reviewed by Ojan Vafai.
Convert json_results_generator.py to output version 4 JSON.
https://bugs.webkit.org/show_bug.cgi?id=60869
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2011-06-18 Berend-Jan Wever <skylined@chromium.org>
Reviewed by Eric Seidel.
Some names with spaces are incorrectly reported as not containing spaces
while running prepare-ChangeLog
https://bugs.webkit.org/show_bug.cgi?id=55572
* Scripts/VCSUtils.pm: Updated CHANGE_LOG_NAME check for spaces to allow
non-alphanumeric characters in names.
2011-06-18 Michael Saboff <msaboff@apple.com>
Reviewed by Eric Seidel.
Enhancement: New Script to Format malloc_history output as tree
https://bugs.webkit.org/show_bug.cgi?id=56378
New script to process malloc_history output to tree format.
* Scripts/malloc-tree: Added.
2011-06-17 Dmitry Lomov <dslomov@google.com>
Reviewed by Daniel Bates.
https://bugs.webkit.org/show_bug.cgi?id=62918
Pass configuration to UnitTests step in master.cfg
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-17 Leandro Pereira <leandro@profusion.mobi>
Reviewed by Eric Seidel.
[EFL] Add Evas-based ImageDiff implementation.
https://bugs.webkit.org/show_bug.cgi?id=61942
Build system changes will be performed as soon as the complete
DumpRenderTree implementation is upstreamed.
* DumpRenderTree/efl/ImageDiff.cpp: Added.
(pixelFromImageData):
(differenceImageFromDifferenceBuffer):
(computeDistanceBetweenPixelComponents):
(calculatePixelDifference):
(calculateDifference):
(getTemporaryFile):
(printImage):
(printImageDifferences):
(resizeEcoreEvasIfNeeded):
(readImageFromStdin):
(parseCommandLineOptions):
(shutdownEfl):
(abortWithErrorMessage):
(initEfl):
(main):
2011-06-17 Tony Chang <tony@chromium.org>
Reviewed by Dirk Pranke.
[chromium] force rebaseline-chromium-webkit-tests to use the chromium port
https://bugs.webkit.org/show_bug.cgi?id=62907
This fixes a bug where the ImageDiff on the user's machine doesn't
match the logic used on the Chromium buildbots. This has the downside
of requiring that the Chromium ImageDiff is built, but it turns out
that even with --tolerance 0, the CG ImageDiff will allow small pixel
differences through.
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-06-17 Brent Fulgham <bfulgham@webkit.org>
[WinCairo] Unreviewed build correction.
* DumpRenderTree/win/DumpRenderTreeCairo.vsprops:
Update WinCairo property sheet to link against the Cairo DLL,
now that we are using a more recent Cairo build.
2011-06-17 Brent Fulgham <bfulgham@webkit.org>
Reviewed by Brian Weinstein.
Create a simple example of using the COM-based C++ listeners.
A new onclick event is attached (via C++) to the WebKit logo
in the default test pattern displayed on startup. Clicking on
the logo causes a message box to be displayed.
https://bugs.webkit.org/show_bug.cgi?id=61885
* WinLauncher/DOMDefaultImpl.h: Added. Stub implementation of
the WebScriptObject and DOMEventListener.
* WinLauncher/WinLauncher.cpp:
(SimpleEventListener::SimpleEventListener): Example implementation
of a simple DOM event listener.
(SimpleEventListener::handleEvent):
(WinLauncherWebHost::didFinishLoadForFrame): Added implementation
to bind a C++ method to the 'onclick' event for the WebKit logo.
(_tWinMain):
* WinLauncher/WinLauncher.h:
* WinLauncher/WinLauncher.vcproj: Add new DOMDefaultImpl.h file.
2011-06-17 Chang Shu <cshu@webkit.org>
Reviewed by Andreas Kling.
[Qt] [WK2] Support undo/redo in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=62809
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):
2011-06-17 Daniel Bates <dbates@rim.com>
Rubber-stamped by Eric Seidel.
Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
https://bugs.webkit.org/show_bug.cgi?id=62767
Part 2 of 3
Update the BuildBot master.cfg to use the new --no-build run-javascriptcore-tests
command line option instead of --skip-build.
We'll remove --skip-build from run-javascriptcore-test in a follow up patch once the
master buildbot picks up the change.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-17 Daniel Bates <dbates@rim.com>
Rubber-stamped by Eric Seidel.
Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
https://bugs.webkit.org/show_bug.cgi?id=62767
Part 1 of 3
Towards renaming the option --skip-build to --no-build. Add --build/--no-build options to
run-javascriptcore-tests. Also, add a remark about these options to the usage message
printed by run-javascriptcore-tests.
We'll remove --skip-build in a follow up patch once the master buildbot picks up the changes.
* Scripts/run-javascriptcore-tests: Additionally, remove unnecessary initialization
of $showHelp.
2011-06-17 Joone Hur <joone.hur@collabora.co.uk>
Reviewed by Martin Robinson.
[GTK] Replace GdkRectangle by cairo_rectangle_int_t
https://bugs.webkit.org/show_bug.cgi?id=60687
Replace GdkRectangle by cairo_rectangle_int_t.
* DumpRenderTree/gtk/PixelDumpSupportGtk.cpp: Replace GdkRectangle by cairo_rectangle_int_t.
(createBitmapContextFromWebView):
* DumpRenderTree/gtk/TextInputController.cpp: Ditto.
(firstRectForCharacterRangeCallback):
2011-06-16 Dmitry Lomov <dslomov@google.com>
Reviewed by David Levin.
https://bugs.webkit.org/show_bug.cgi?id=62603
Run run-api-tests on bots in debug mode.
Release mode is disabled due to https://bugs.webkit.org/show_bug.cgi?id=61812.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-16 Leandro Pereira <leandro@profusion.mobi>
Reviewed by Eric Seidel.
[EFL] DumpRenderTree: Add GCController, PixelDumpSupport and
WorkQueueItem.
They will be plugged into the build system once the EFL's DRT is
fully upstreamed.
https://bugs.webkit.org/show_bug.cgi?id=61958
* DumpRenderTree/efl: Added.
* DumpRenderTree/efl/GCControllerEfl.cpp: Added.
(GCController::collect):
(GCController::collectOnAlternateThread):
(GCController::getJSObjectCount):
* DumpRenderTree/efl/PixelDumpSupportEfl.cpp: Added.
(createBitmapContextFromWebView):
* DumpRenderTree/efl/WorkQueueItemEfl.cpp: Added.
(LoadItem::invoke):
(LoadHTMLStringItem::invoke):
(ReloadItem::invoke):
(ScriptItem::invoke):
(BackForwardItem::invoke):
* DumpRenderTree/efl/WorkQueueItemEfl.h: Added.
(WorkQueueItem::~WorkQueueItem):
(LoadItem::LoadItem):
(LoadHTMLStringItem::LoadHTMLStringItem):
(ScriptItem::ScriptItem):
(LoadingScriptItem::LoadingScriptItem):
(LoadingScriptItem::invoke):
(NonLoadingScriptItem::NonLoadingScriptItem):
(NonLoadingScriptItem::invoke):
(BackForwardItem::BackForwardItem):
(BackItem::BackItem):
(ForwardItem::ForwardItem):
2011-06-16 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix, make sure the waf build checks the webKitBranchBuild
preference like the perl scripts do.
* waf/build/build_utils.py:
* waf/build/settings.py:
2011-06-16 Tommy Widenflycht <tommyw@google.com>
Reviewed by Tony Gentilcore.
MediaStream API: Runtime enabled features
https://bugs.webkit.org/show_bug.cgi?id=62798
Add the WebRuntimeFeatures API for Media Streams and enable the feature by default in DumpRenderTree.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
2011-06-16 Torsten Kurbad <gentoo@tk-webart.de>
Reviewed by Eric Seidel.
Add support for sys.platform == 'linux3' by relaxing
conditional statements and/or extending lookup dicts.
This way, Linux kernels >=3.0-rc1 get supported in the
build process.
https://bugs.webkit.org/show_bug.cgi?id=62730
* Scripts/webkitpy/common/system/executive.py:
* Scripts/webkitpy/common/system/file_lock.py:
* Scripts/webkitpy/common/system/path_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
* Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
* Scripts/webkitpy/layout_tests/port/factory.py:
* Scripts/webkitpy/layout_tests/port/factory_unittest.py:
* Scripts/webkitpy/layout_tests/port/qt.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-06-16 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r88796.
http://trac.webkit.org/changeset/88796
https://bugs.webkit.org/show_bug.cgi?id=62790
It made fast/dom/nodesFromRect-basic.html time out on Qt,
64-bit, debug mode (Requested by Ossy on #webkit).
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::nodesFromRect):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/qt/PlainTextControllerQt.cpp:
(PlainTextController::plainText):
* DumpRenderTree/qt/PlainTextControllerQt.h:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Unreviewed, build fix.
Revert the patches in r88995 and r88999; the http servers don't
seem to be shutting down reliably.
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Unreviewed, build fix.
Add some debugging messages to triage problems introduced in
r88995.
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: should clean up stale server processes from a previous run
https://bugs.webkit.org/show_bug.cgi?id=62615
This change fixes cases where NRWT might start a server and then
exit, leaving the server still running. Now NRWT should detect
this case and clean up on the next run, without shutting down
any processes it didn't start.
This change also fixes a bug in executive.kill_process on UNIX
where zombies weren't being handled properly.
* Scripts/webkitpy/common/system/executive.py:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: fix http, websocket server startup, shutdown
https://bugs.webkit.org/show_bug.cgi?id=62180
This change refactors all of the server-related code to
have consistent semantics. It pushes the start()/stop()
logic into the base class and adds four much smaller routines
for specialization:
_prepare_config() - for optionally writing config files etc.
to disk prior to starting the server
_remove_stale_log_files()
_spawn_process() - the actual launch of the subprocess
_cleanup_after_stop() - and custom cleanup code
Prior to this change the three server implementations were
fairly inconsistent, leading to lots of weirdness on the bots
with servers being left around, etc.
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
webkitpy: clean up code prior to functional changes for server startup/shutdown
https://bugs.webkit.org/show_bug.cgi?id=62256
This patch does a bunch of minor cleanup of the code, but there
should be no functional changes except that all of the modules
will now use the same exception type. Other changes include
removing unused code paths and functions, pushing 'name' and
'pid_file' into the base class of the server objects, and using
the Filesystem object instead of the codecs, tempfile, and
shutils modules.
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
webkitpy: add integration tests for new-run-webkit-httpd, stop calling shut_down_http_server
https://bugs.webkit.org/show_bug.cgi?id=62251
shut_down_http_server() was a total hack that was only used by
new-run-webkit-httpd, so I've moved the code there and switched
to using executive.kill_process() for the common case. The
method itself will be removed in the patch on bug 59993.
* Scripts/new-run-webkit-httpd:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Unreviewed, build fix.
Fix test-webkitpy failure introduced in r88986.
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-06-15 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
NRWT: doesn't support webarchives, has stale skipping logic.
https://bugs.webkit.org/show_bug.cgi?id=58293
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-06-15 Jeremy Orlow <jorlow@webkit.org>
Change my email to @webkit.org
* Scripts/webkitpy/common/config/committers.py:
2011-06-15 Nikolas Zimmermann <nzimmermann@rim.com>
Reviewed by David Hyatt.
CSS 2.1: Support replaced elements with relative intrinsic sizes
https://bugs.webkit.org/show_bug.cgi?id=15849
CSS 2.1 failure: html4/replaced-intrinsic-ratio-001
https://bugs.webkit.org/show_bug.cgi?id=53099
Recognize *.htm tests in the old & new DRT implementation, needed for the vanilla css2.1 html4 tests.
* Scripts/old-run-webkit-tests:
* Scripts/webkitpy/layout_tests/port/test_files.py:
2011-06-15 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Add a statusbar to MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=62634
It shows the url of the current hovered link.
* MiniBrowser/gtk/BrowserWindow.c:
(browserWindowConstructed): Create the GtkSatusbar.
(mouseDidMoveOverElement): Update statusbar text.
(browserWindowUIClientInit): Add implementation for
mouseDidMoveOverElement().
* MiniBrowser/gtk/GNUmakefile.am: Add new files to compilation.
* MiniBrowser/gtk/WebBundle/WebBundleMain.c: Added.
(mouseDidMoveOverElement): Pass the url of the current hovered
link to the UI process.
(didCreatePage): Set the UI client adding an implementation for
mouseDidMoveOverElement().
(WKBundleInitialize):
* MiniBrowser/gtk/main.c:
(createWKContextWithInjectedBundle): Create a global context with
the injected bundle.
(loadURI): Use the global context.
(main):
2011-06-14 Lucas Forschler <lforschler@apple.com>
Reviewed by Stephanie Lewis.
https://bugs.webkit.org/show_bug.cgi?id=62495
Bug 62495 - combine windows and mac kill-old-processes script
Add a platform agnostic script. Remove the recently added mac flavor.
Note: Leaving the windows one alone, since it is currently used by the WebKit bots.
* BuildSlaveSupport/kill-old-processes: Added.
* BuildSlaveSupport/mac: Removed.
* BuildSlaveSupport/mac/kill-old-processes: Removed.
2011-06-14 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: handle worker exceptions cleanly
https://bugs.webkit.org/show_bug.cgi?id=62614
This change modifiers new-run-webkit-tests to handle exceptions
raised by worker threads better, by:
- capturing the worker's stack and logging it in the manager
- propagating the worker's exception in the caller correctly
- attempting to cancel the workers and clean up even when
we get an unexpected exception
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2011-06-14 Qi Zhang <qi.2.zhang@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] fast/dom/HTMLScriptElement/nested-execution.html failed
https://bugs.webkit.org/show_bug.cgi?id=62227
QtWebkit does not yet support different CacheModels. This change will
expose setCacheModel() with a stub implementation, which is enough to pass the LayoutTest.
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setCacheModel):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-06-14 Caio Marcelo de Oliveira Filho <caio.oliveira@openbossa.org>
Reviewed by Andreas Kling.
[Qt] tst_QWebFrame::overloadedSlots() fails
https://bugs.webkit.org/show_bug.cgi?id=37319
Since we don't implictly convert 'document' object to QWebElement
in metacalls anymore, change the controller to expect QVariantMap instead.
The method plainText() was updated to use QVariantMap as well to let the bridge
do the conversion directly for us.
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::nodesFromRect):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/qt/PlainTextControllerQt.cpp:
(PlainTextController::plainText):
* DumpRenderTree/qt/PlainTextControllerQt.h:
2011-06-14 Andras Becsi <abecsi@webkit.org>
Reviewed by Csaba Osztrogonác.
Remove dead code in DumpRenderTree/TestNetscapePlugIn/main.cpp
https://bugs.webkit.org/show_bug.cgi?id=62630
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(NPP_New): remove dead variable 'forceCarbon'
2011-06-13 Hayato Ito <hayato@chromium.org>
Reviewed by Ojan Vafai.
[NRWT] Fix a WebKitDriver so that it returns an actual hash value if DRT returns that.
https://bugs.webkit.org/show_bug.cgi?id=59188
A current WebKitDriver doesn't try to return an actual hash value if DRT does not return an image.
WebKitDriver should always return an actual hash value if DRT returns that.
No tests since I'll add sample reftests to make sure mismatch reftests work later,
which should be in a separate patch.
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-06-13 Kent Tamura <tkent@chromium.org>
[Chromium-win] Fix two wrong assertions exposed by r88757.
* DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
(WebThemeEngineDRTWin::paintTrackbar):
2011-06-13 Alexey Proskuryakov <ap@apple.com>
Added an updated e-mail address for Matt Lilek, to make Bugzilla autocomplete
and commit queue happy.
* Scripts/webkitpy/common/config/committers.py:
2011-06-13 Chris Rogers <crogers@google.com>
Unreviewed build fix.
Build fix to chromium DRT until we can rebaseline some tests
https://bugs.webkit.org/show_bug.cgi?id=62608
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
2011-06-13 Chris Rogers <crogers@google.com>
Reviewed by Dirk Pranke.
Add Web Audio support to chromium DRT
https://bugs.webkit.org/show_bug.cgi?id=62306
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::reset):
(LayoutTestController::setEncodedAudioData):
* DumpRenderTree/chromium/LayoutTestController.h:
(LayoutTestController::encodedAudioData):
(LayoutTestController::shouldDumpAsAudio):
(LayoutTestController::setShouldDumpAsAudio):
* DumpRenderTree/chromium/TestEventPrinter.cpp:
(DRTPrinter::handleAudioHeader):
(TestShellPrinter::handleAudioHeader):
* DumpRenderTree/chromium/TestEventPrinter.h:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
(TestShell::dump):
2011-06-13 Jaehun Lim <ljaehun.lim@samsung.com>
Reviewed by Eric Seidel.
[EFL] Add load error handler to EWebLauncher
https://bugs.webkit.org/show_bug.cgi?id=62347
Add a handler function of "load,error" to EWebLauncher.
EWebLauncher now displays a simple error page on load errors.
* EWebLauncher/main.c:
(on_load_error):
(browserCreate):
2011-06-13 Martin Robinson <mrobinson@igalia.com>
Reviewed by Eric Seidel.
[GTK] [Qt] Eliminate duplicate TestNetscapePlugin implementation
https://bugs.webkit.org/show_bug.cgi?id=62385
Remove duplicate TestNetscapePlugIn implementation. Having two copies
of this code makes keeping it in sync much more difficult. The files are
almost identical and this change ports the X11 changes to main.cpp.
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(NPP_New): Do not set the plugin as windowless always. Doing this breaks
some results on Qt and this seems to be unnecessary for the moment.
(keyEventToChar): Added this helper which converts a X11 keycode into a char.
(handleEventX11): Use the handler to properly convert the keycode. Do not print
adjustCursorEvent output. The tests currently do not trigger this for Cocoa and thus
it should not be in the results.
(NPP_GetValue): Properly handle NPPVpluginNeedsXEmbed and make it clearer why the
XP_UNIX block is first by guarding against using a null instance.
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Update source list add
the XP_UNIX define for Linux.
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Removed.
* GNUmakefile.am: Update source list.
2011-06-13 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r88671.
http://trac.webkit.org/changeset/88671
https://bugs.webkit.org/show_bug.cgi?id=62591
"didn't work" (Requested by dpranke on #webkit).
* Scripts/new-run-webkit-httpd:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-13 Nate Chapin <japhet@chromium.org>
Reviewed by Darin Fisher.
New test method on TestNetscapePlugin for https://bugs.webkit.org/show_bug.cgi?id=61482.
* DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
2011-06-13 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
webkitpy: add integration tests for new-run-webkit-httpd, stop calling shut_down_http_server
https://bugs.webkit.org/show_bug.cgi?id=62251
shut_down_http_server() was a total hack that was only used by
new-run-webkit-httpd, so I've moved the code there and switched
to using executive.kill_process() for the common case. The
method itself will be removed in the patch on bug 59993.
* Scripts/new-run-webkit-httpd:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-06-13 Tony Chang <tony@chromium.org>
Reviewed by Dimitri Glazkov.
rename ENABLE_NEW_FLEXBOX to ENABLE_CSS3_FLEXBOX
https://bugs.webkit.org/show_bug.cgi?id=62578
* Scripts/build-webkit:
2011-06-13 Tony Chang <tony@chromium.org>
Reviewed by Adam Barth.
rename ENABLE_FLEXBOX to ENABLE_NEW_FLEXBOX
https://bugs.webkit.org/show_bug.cgi?id=62545
* Scripts/build-webkit:
2011-06-13 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Export an API similar to WebKit1
https://bugs.webkit.org/show_bug.cgi?id=57820
Build GtkLauncher2 for WebKit2. It's actually the same GtkLauncher
code, but linking to libwebkit2gtk.
* GNUmakefile.am:
* GtkLauncher/main.c:
(createBrowser):
(main):
2011-06-12 MORITA Hajime <morrita@google.com>
Unreviewed, rolling out r88625.
http://trac.webkit.org/changeset/88625
https://bugs.webkit.org/show_bug.cgi?id=61073
Breaks SL Webkit2 Tests
* WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
* WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
* WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::didClearWindowForFrame):
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2011-06-12 Mahesh Kulkarni <mahesh.kulkarni@nokia.com>
Reviewed by Antonio Gomes.
[Qt] LayoutTestController needs setTextDirection implementation
https://bugs.webkit.org/show_bug.cgi?id=62442
Implements LayoutController.setTextDirection.
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setTextDirection):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-06-12 Dominic Cooney <dominicc@chromium.org>
Reviewed by Hajime Morita.
Add window.internals to WebKit2's WebKitTestRunner.
https://bugs.webkit.org/show_bug.cgi?id=61073
Test: fast/harness/internals-object.html
* WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
* WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
* WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::didClearWindowForFrame): initialize window.internals
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2011-06-12 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r88616.
http://trac.webkit.org/changeset/88616
https://bugs.webkit.org/show_bug.cgi?id=62517
It broke editing/text-iterator/findString.html (Requested by
tonikitoo on #webkit).
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-06-12 Mahesh Kulkarni <mahesh.kulkarni@nokia.com>
Reviewed by Antonio Gomes.
[Qt] LayoutTestController needs setTextDirection implementation
https://bugs.webkit.org/show_bug.cgi?id=62442
Implements LayoutController.setTextDirection.
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setTextDirection):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-06-10 Ryosuke Niwa <rniwa@webkit.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: 'images' links on results.html don't work
https://bugs.webkit.org/show_bug.cgi?id=62194
Use relative path when generating images-diffs.html
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
Added _output_testname.
2011-06-10 Lucas Forschler <lforschler@apple.com>
Reviewed by Stephanie Lewis.
Script to kill old processes on Mac.
This matches the windows version.
https://bugs.webkit.org/show_bug.cgi?id=62479
* BuildSlaveSupport/mac: Added.
* BuildSlaveSupport/mac/kill-old-processes: Added.
2011-06-10 Adam Barth <abarth@webkit.org>
Reviewed by James Robinson.
When sheriffbot reports failures on IRC instead of saying "(and more..)" it should say the number of tests that are failing
https://bugs.webkit.org/show_bug.cgi?id=62489
* Scripts/webkitpy/tool/bot/sheriff.py:
* Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
2011-06-10 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
add a compile guard ENABLE(FLEXBOX)
https://bugs.webkit.org/show_bug.cgi?id=62049
* Scripts/build-webkit:
2011-06-10 Mark Rowe <mrowe@apple.com>
Reviewed by Dimitri Glazkov.
<rdar://problem/9562114> Fix DumpRenderTree build for production configuration.
Ensure that DumpRenderTree can find libWebCoreTestSupport.dylib and the associated
header file.
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/mac/Configurations/Base.xcconfig:
2011-06-09 Csaba Osztrogonác <ossy@webkit.org>
Unreviewed rolling out r88471, because it broke plugin tests on Qt.
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(handleEventX11):
(NPP_GetValue):
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Added.
(webkit_test_plugin_new_instance):
(webkit_test_plugin_destroy_instance):
(webkit_test_plugin_set_window):
(executeScript):
(webkit_test_plugin_new_stream):
(webkit_test_plugin_destroy_stream):
(webkit_test_plugin_stream_as_file):
(webkit_test_plugin_write_ready):
(webkit_test_plugin_write):
(webkit_test_plugin_print):
(keyEventToChar):
(webkit_test_plugin_handle_event):
(webkit_test_plugin_url_notify):
(webkit_test_plugin_get_value):
(webkit_test_plugin_set_value):
(NP_GetMIMEDescription):
(NP_Initialize):
(NP_Shutdown):
(NP_GetValue):
* GNUmakefile.am:
2011-06-09 Martin Robinson <mrobinson@igalia.com>
Reviewed by Eric Seidel.
[GTK] Implement the next bit of the TextInputController (hasMarkedText and markedRange)
https://bugs.webkit.org/show_bug.cgi?id=55603
Add callbacks to support hasMarkedText and markedRange and also clean up existing
TextInputController support.
* DumpRenderTree/gtk/TextInputController.cpp:
(setMarkedTextCallback): Accept parameters in terms of start and length, instead
of start and end. This matches the other ports. Do not use GLib style early returns.
(hasMarkedTextCallback): Added.
(markedRangeCallback): Added.
(insertTextCallback): Cleanup as described above.
(unmarkTextCallback): Cleanup as described above.
(firstRectForCharacterRangeCallback): Cleanup as described above.
(selectedRangeCallback): Cleanup as described above.
2011-06-09 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: support webaudio in chromium driver
https://bugs.webkit.org/show_bug.cgi?id=62226
* Scripts/webkitpy/layout_tests/port/chromium.py:
2011-06-09 Dan Bernstein <mitz@apple.com>
Reviewed by Anders Carlsson.
Added a test for Vector::reverse()
https://bugs.webkit.org/show_bug.cgi?id=62393
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WTF/VectorReverse.cpp: Added.
(TestWebKitAPI::TEST):
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
2011-06-09 Martin Robinson <mrobinson@igalia.com>
Reviewed by Andreas Kling.
[GTK] [Qt] Eliminate duplicate TestNetscapePlugin implementation
https://bugs.webkit.org/show_bug.cgi?id=62385
Remove duplicate TestNetscapePlugIn implementation. Having two copies
of this code makes keeping it in sync much more difficult. The files are
almost identical and this change ports the X11 changes to main.cpp.
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(keyEventToChar): Added this helper which converts a X11 keycode into a char.
(handleEventX11): Use the handler to properly convert the keycode. Do not print
adjustCursorEvent output. The tests currently do not trigger this for Cocoa and thus
it should not be in the results.
(NPP_GetValue): Properly handle NPPVpluginNeedsXEmbed.
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Update source list.
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Removed.
* GNUmakefile.am: Update source list.
2011-06-09 Noel Gordon <noel.gordon@gmail.com>
Reviewed by Tony Chang.
Test plugin should support event logging on the windows port.
https://bugs.webkit.org/show_bug.cgi?id=61721
The WebKit win port interprets a 0 return from plugin NPP_HandleEvent()
calls to mean the plugin handled the event.
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(handleEventWin): Return 0 for each event on windows.
2011-06-09 Chang Shu <cshu@webkit.org>
Reviewed by Andreas Kling.
[Qt] [WK2] Many editing tests failed on missing shouldBeginEditingInDOMRange:range printout
https://bugs.webkit.org/show_bug.cgi?id=62381
Activate QGraphicsScene and setFocus on QGraphicsItem to make sure the view has the focus.
QWKPagePrivate::isViewFocused() will check it later.
* WebKitTestRunner/qt/PlatformWebViewQt.cpp:
(WTR::PlatformWebView::PlatformWebView):
2011-06-09 Robert Hogan <robert@webkit.org>
Reviewed by Andreas Kling.
Teach Qt about window.internals
https://bugs.webkit.org/show_bug.cgi?id=61074
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::initJSObjects):
2011-06-09 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Disable running layout tests on Mac EWS bots (and fix cr-linux bots to run them again)
https://bugs.webkit.org/show_bug.cgi?id=62362
The mac bots are failing inspector tests. We don't know why yet.
So for now, turning run-webkit-tests off for the mac-ews.
Also noticed that due to the wrong variable name the cr-linux
bots weren't running tests!
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2011-06-09 Zoltan Herczeg <zherczeg@inf.u-szeged.hu>
Reviewed by Csaba Osztrogonac.
[Qt][WK2] Canvas tests make css2.1 ones failing
https://bugs.webkit.org/show_bug.cgi?id=53427
Adding a workaround for this issue by mimicing the behaviour of DumpRenderTree.
Since the Qt port should eventually go to QRawFont, this issue will be resolved
in the future. Both clearMemoryCaches() and removeAllApplicationFonts() are
needed for the workaround.
* WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
(WTR::LayoutTestController::platformInitialize):
2011-06-09 Mario Sanchez Prada <msanchez@igalia.com>
Reviewed by Chris Fleizach.
[GTK] Provide a way in DRT to check the platform name
https://bugs.webkit.org/show_bug.cgi?id=56855
Patch for adding 'platformName' static property to DRT's
LayoutController and provide platform specific implementations.
In principle, this feature will be only of interest for writing
accessibility tests, so that's why it's only implemented in the
Mac, Gtk and Win platforms.
* DumpRenderTree/LayoutTestController.h: New method platformName().
* DumpRenderTree/LayoutTestController.cpp:
(getPlatformNameCallback): New callback for 'platformName'.
(LayoutTestController::staticValues): New static property.
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::platformName): Implemented.
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::platformName): Implemented.
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::platformName): Implemented.
2011-06-08 Mike Lawther <mikelawther@chromium.org>
Reviewed by James Robinson.
Add mikelawther to committers.py
https://bugs.webkit.org/show_bug.cgi?id=62354
* Scripts/webkitpy/common/config/committers.py:
2011-06-08 Kent Tamura <tkent@chromium.org>
[Chromium] Fix a typo.
* DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
(WebThemeControlDRTWin::draw):
2011-06-08 Kent Tamura <tkent@chromium.org>
Reviewed by Hajime Morita.
[Chromium] Support vertical sliders in Windows DRT
https://bugs.webkit.org/show_bug.cgi?id=62352
* DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
(WebThemeControlDRTWin::draw):
Paint VerticalSliderThumbType and VerticalSliderTrackType.
* DumpRenderTree/chromium/WebThemeControlDRTWin.h:
Add VerticalSliderTrackType and VerticalSliderThumbType.
* DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
(WebThemeEngineDRTWin::paintTrackbar):
Convert Windows identifiers to the internal identifiers for vertical sliders.
2011-06-08 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
add a short-flag for --experimental-fully-parallel
https://bugs.webkit.org/show_bug.cgi?id=62321
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-06-08 Ryosuke Niwa <rniwa@webkit.org>
Reviewed by Tony Chang.
new-run-webkit-tests: don't turn the tree red when there are new passes
https://bugs.webkit.org/show_bug.cgi?id=62311
Look for "new passes" instead of "passed unexpectedly" in evaluateCommand
to match commandComplete.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-08 Tom Hudson <tomhudson@google.com>
Reviewed by Mihai Parparita.
Allow sorting in RebaselineServer based on 'metric' field in unexpected_results.json
https://bugs.webkit.org/show_bug.cgi?id=60964
* Scripts/webkitpy/tool/commands/data/rebaselineserver/index.html:
Add 'Sort tests by metric' link.
* Scripts/webkitpy/tool/commands/data/rebaselineserver/main.js:
(disableSorting): Activate 'Sort tests by metric' link.
(enableSorting): Deactivate 'Sort tests by metric' link.
(selectDirectory): Call enableSorting()/disableSorting() depending
on currently selected failure type, and sort tests if requested.
* Scripts/webkitpy/tool/commands/data/rebaselineserver/main.css:
New .disabled-control class for deactivated links.
2011-06-08 Tom Sepez <tsepez@chromium.org>
Reviewed by Adam Barth.
Add web permissions callbacks for mixed content blocking.
https://bugs.webkit.org/show_bug.cgi?id=61946
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::setAllowDisplayOfInsecureContent):
(LayoutTestController::setAllowRunningOfInsecureContent):
(LayoutTestController::overridePreference):
* DumpRenderTree/chromium/WebPermissions.cpp:
(WebPermissions::allowDisplayingInsecureContent):
(WebPermissions::allowRunningInsecureContent):
(WebPermissions::setDisplayingInsecureContentAllowed):
(WebPermissions::setRunningInsecureContentAllowed):
(WebPermissions::reset):
* DumpRenderTree/chromium/WebPermissions.h:
2011-06-08 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
webkitpy: fix typo causing us to skip the port unit tests in chromium_linux
https://bugs.webkit.org/show_bug.cgi?id=62259
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
2011-06-08 Ryosuke Niwa <rniwa@webkit.org>
Reviewed by Tony Chang.
new-run-webkit-tests: master.cfg errors when nrwt successfully runs test
https://bugs.webkit.org/show_bug.cgi?id=62303
Access self.incorrectLayoutLines instead of incorrectLayoutLines in evaluateCommand.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-08 Greg Simon <gregsimon@chromium.org>
Reviewed by Dimitri Glazkov.
Control Indexeddb backends from LayoutTestController
https://bugs.webkit.org/show_bug.cgi?id=61000
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setOverrideIndexedDBBackingStore):
(LayoutTestController::clearAllDatabases):
* DumpRenderTree/chromium/LayoutTestController.h:
2011-06-08 Andreas Kling <kling@webkit.org>
Add Pierre Rossi to contributors list for EZ-CC.
* Scripts/webkitpy/common/config/committers.py:
2011-06-08 Ilya Sherman <isherman@chromium.org>
Reviewed by Andreas Kling.
Add Ilya Sherman to list of committers
https://bugs.webkit.org/show_bug.cgi?id=62270
* Scripts/webkitpy/common/config/committers.py:
2011-06-07 Ryosuke Niwa <rniwa@webkit.org>
Reviewed by Tony Chang.
new-run-webkit-tests: Bot master should print useful information on waterfall/console for nrwt
https://bugs.webkit.org/show_bug.cgi?id=62178
Added commandComplete and evaluateCommand for NewRunWebKitTests class
to parse new-run-webkit-tests' results.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-06-07 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
webkitpy: add a popen() call to executive
https://bugs.webkit.org/show_bug.cgi?id=62179
This change adds a popen() wrapper call to the Executive object.
This will allow other webkitpy scripts that currently call
subprocess directly to use Executive, instead.
No additional tests are necessary (the existing tests cover the code).
* Scripts/webkitpy/common/system/executive.py:
2011-06-07 Sam Weinig <sam@webkit.org>
Reviewed by Anders Carlsson.
Remove repeated IPC traffic from SetRenderTreeSize message
https://bugs.webkit.org/show_bug.cgi?id=62244
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage):
Add stub for didLayoutForFrame.
2011-06-07 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
rebaseline-chromium-webkit-tests: fix baselining order for linux, linux_x86_64
https://bugs.webkit.org/show_bug.cgi?id=62241
* Scripts/webkitpy/layout_tests/port/chromium.py:
2011-06-07 Adam Barth <abarth@webkit.org>
Rubber-stamped by Eric Seidel.
Enable tests on the mac-ews
When we bring the mac-ews back up, we're going to try running tests!
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2011-06-07 Dmitry Lomov <dslomov@google.com>
Reviewed by David Levin.
https://bugs.webkit.org/show_bug.cgi?id=62215
Allow comparisons with 0 in LIKELY and UNLIKELY macros.
* Scripts/webkitpy/style/checkers/cpp.py:
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
2011-06-07 Nico Weber <thakis@chromium.org>
Reviewed by Dimitri Glazkov.
[chromium] -Wdelete-non-virtual-dtor pass for DumpRenderTree
https://bugs.webkit.org/show_bug.cgi?id=62210
The change to TestEventPrinter fixes a latent bug, because
objects are deleted through the TestEventPrinter type, but none
of the subclasses have destructors or non-POD members.
The changes to NotificationPresenter and WebViewHost do _not_ to fix a
real bug, they just make clang's -Wdelete-non-virtual-dtor happy. As
discussed at http://codereview.chromium.org/7094005/, we prefer making
leaf class destructors virtual over making the leaf classes final.
* DumpRenderTree/chromium/NotificationPresenter.cpp:
(NotificationPresenter::~NotificationPresenter):
* DumpRenderTree/chromium/NotificationPresenter.h:
* DumpRenderTree/chromium/TestEventPrinter.cpp:
(TestEventPrinter::~TestEventPrinter):
* DumpRenderTree/chromium/TestEventPrinter.h:
* DumpRenderTree/chromium/WebViewHost.h:
2011-06-06 Ryosuke Niwa <rniwa@webkit.org>
Reviewed by Dirk Pranke.
[chromium] Chromium bots on build.webkit.org don't upload results.html
https://bugs.webkit.org/show_bug.cgi?id=61961
Don't fix the path on chromium port erroneously.
* BuildSlaveSupport/test-result-archive:
2011-06-06 Martin Robinson <mrobinson@igalia.com>
Fix the GTK+ build by ensuring that the autogenerated sources
necessary for libWebCoreInternals are recorded in a predeclared
variable before assigning to BUILT_SOURCES.
* GNUmakefile.am: Predeclare libwebcoreinternals_built_sources.
2011-06-06 Yong Li <yoli@rim.com>
Reviewed by Eric Seidel.
https://bugs.webkit.org/show_bug.cgi?id=62097
Fix the problem that integral bitfield check incorrectly barked at ?: expressions.
* Scripts/webkitpy/style/checkers/cpp.py:
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
2011-06-06 Xan Lopez <xlopez@igalia.com>
Reviewed by Martin Robinson.
[GTK] Enable feature/symbol detection in NWRT/GTK
https://bugs.webkit.org/show_bug.cgi?id=62136
Disable feature detection through DRT, since we don't support it,
and set the libwebcore patch correctly so that symbol detection
through 'nm' works.
* Scripts/webkitpy/layout_tests/port/gtk.py:
2011-06-06 Shishir Agrawal <shishir@chromium.org>
Reviewed by Tony Gentilcore.
Add a "prerender" state for page visibility.
https://bugs.webkit.org/show_bug.cgi?id=62062
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::setPageVisibility):
2011-06-06 Csaba Osztrogonác <ossy@webkit.org>
[Qt] Implement download feature for QtTestBrowser
https://bugs.webkit.org/show_bug.cgi?id=61865
Buildfix for --minimal and CONFIG+=qt_minimal build after r88161.
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::LauncherWindow):
* QtTestBrowser/launcherwindow.h:
2011-06-06 Barát Tibor <Barat.Tibor@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[Qt] Implement download feature for QtTestBrowser
https://bugs.webkit.org/show_bug.cgi?id=61865
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::LauncherWindow):
(LauncherWindow::downloadRequest):
(LauncherWindow::fileDownloadFinished):
* QtTestBrowser/launcherwindow.h:
2011-06-04 Dominic Cooney <dominicc@chromium.org>
Reviewed by Dimitri Glazkov.
Make DumpRenderTree initialize window.internals.
https://bugs.webkit.org/show_bug.cgi?id=61076
* DumpRenderTree/win/DumpRenderTreeCommon.vsprops: link WebCoreTestSupport.
* DumpRenderTree/win/FrameLoadDelegate.cpp:
(FrameLoadDelegate::didClearWindowObjectForFrameInStandardWorld):
2011-06-06 Alexandru Chiculita <achicu@adobe.com>
Reviewed by Kent Tamura.
Add ENABLE_CSS_EXCLUSIONS support for build-webkit script
https://bugs.webkit.org/show_bug.cgi?id=61628
* Scripts/build-webkit:
2011-06-06 Mihnea Ovidenie <mihnea@adobe.com>
Reviewed by Kent Tamura.
Add ENABLE(CSS_REGIONS) guard for CSS Regions support
https://bugs.webkit.org/show_bug.cgi?id=61631
* Scripts/build-webkit:
2011-06-06 Carlos Garcia Campos <cgarcia@igalia.com>
Unreviewed. Fix the GTK+ build.
* GNUmakefile.am: Use javascriptcore_cflags when building libWebCoreInternals.
2011-06-05 Alexey Proskuryakov <ap@apple.com>
Add an e-mail address for Cameron McCormack that matches Bugzilla (so that auto-completion
works properly).
* Scripts/webkitpy/common/config/committers.py:
2011-06-04 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix. Update git branch build dir handling to match what other
WebKit scripts expect, so other scripts work properly with wx.
* waf/build/build_utils.py:
* waf/build/settings.py:
2011-06-04 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix. Restore the PPC build and allow users to specify architectures
to build on Mac.
* waf/build/settings.py:
* wx/install-unix-extras:
2011-06-04 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fixes for recent trunk changes.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::setTextDirection):
* waf/build/settings.py:
2011-06-04 Martin Robinson <mrobinson@igalia.com>
Fix the GTK+ build.
* GNUmakefile.am: Add some missing newline escape.
2011-06-03 Martin Robinson <mrobinson@igalia.com>
Try to fix the GTK+ bulid.
* GNUmakefile.am: Add JSDOMWrapper.cpp to the libWebCoreInternals
source list.
2011-06-03 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
nrwt: should skip chromium test expectation overrides on upstream bots
https://bugs.webkit.org/show_bug.cgi?id=61888
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2011-06-03 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: minor logging, code cleanup
https://bugs.webkit.org/show_bug.cgi?id=62055
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-06-03 Martin Robinson <mrobinson@igalia.com>
Try to fix the build failure on the GTK+ 32-bit Release bot.
* GNUmakefile.am: Add JSInternal.{cpp,h} to the BUILT_SOURCES list.
2011-05-31 Martin Robinson <mrobinson@igalia.com>
Reviewed by Ryosuke Niwa.
[GTK] Support smart replace for the pasteboard
https://bugs.webkit.org/show_bug.cgi?id=61734
Call into the DumpRenderTreeSupportGtk hooks to toggle smart replace.
This is required to produce consistent test results.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues): Set the smart replace setting to on by default.
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::setSmartInsertDeleteEnabled): Added an implementation of this
methods which calls into DumpRenderTreeSupportGtk.
2011-06-03 Martin Robinson <mrobinson@igalia.com>
Reviewed by Dimitri Glazkov.
Teach Gtk build about window.internals
https://bugs.webkit.org/show_bug.cgi?id=61071
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(webViewWindowObjectCleared): Inject the new internals object when the
window object clears.
* GNUmakefile.am: Add new source files and includes necessary for window.internals.
Also do a little bit of cleanup.
2011-06-03 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
Worker may not be stopped after LT.
https://bugs.webkit.org/show_bug.cgi?id=61847
Re-submit a variant of the first fix for this bug. It's possible
that actually calling cleanup() in the worker threads was taking
too long and we triggered the assertions in the manager thread.
I've bumped up the timeout in the manager and added better
diagnostics. I also cleaned up the cleanup() code a bit because
there was some unnecessary if-checks.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-06-03 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: fix duplicate logging regression
https://bugs.webkit.org/show_bug.cgi?id=61981
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
2011-06-03 Robert Kroeger <rjkroege@chromium.org>
Reviewed by Adam Barth.
Modified addTouchPoint to correctly re-use touchPoint id values starting
from the lowest not-un-use touchPoint id as is done by the embedder
APIs. (Fixes BUGCR36415)
https://bugs.webkit.org/show_bug.cgi?id=62016
* DumpRenderTree/chromium/EventSender.cpp:
(EventSender::addTouchPoint):
2011-06-03 Cary Clark <caryclark@google.com>
Reviewed by Eric Seidel.
Draw DRT scroll bar thumb with CG when WebKit renders with Skia
https://bugs.webkit.org/show_bug.cgi?id=61534
Skia on Mac uses Skia to render WebKit, and CG
to render UI elements. The CG elements need a
transcribed graphics context that preserves the
canvas matrix, and the canvas clip.
The SkiaBitLocker utility class sets up a CGContext
from the SkCanvas, locks the bitmap's bits, and
releases the lock when the class goes out of scope.
The SkiaBitLocker implementation is described by
this patch:
http://codereview.chromium.org/7031006/
When WebKit uses Skia for rendering, the WebCanvas
is an SkCanvas. Create an equivalent CGContext to
draw the scrollbar thumb.
As the Chrome Mac port does not yet use Skia, this
patch has no effect and requires no tests.
* DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
(WebThemeEngineDRTMac::paintHIThemeScrollbarThumb):
2011-06-03 Andras Becsi <abecsi@webkit.org>
Unreviewed build fix.
Fix typo.
* Scripts/old-run-webkit-tests:
2011-06-03 Andras Becsi <abecsi@webkit.org>
Reviewed by Csaba Osztrogonác.
Print error message when there is no httpd present on the system
https://bugs.webkit.org/show_bug.cgi?id=61939
* Scripts/old-run-webkit-tests: Propose to run with --no-httpd.
* Scripts/webkitperl/httpd.pm:
2011-06-03 Mikhail Naganov <mnaganov@chromium.org>
Reviewed by Yury Semikhatsky.
Web Inspector: [Chromium] cpu-profiler-profiling layout test is flaky on Linux Debug.
https://bugs.webkit.org/show_bug.cgi?id=61533
Skip cpu-profiler-profiling in debug, implement a fast headless alternative.
* DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
(DRTDevToolsAgent::setJavaScriptProfilingEnabled):
* DumpRenderTree/chromium/DRTDevToolsAgent.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setJavaScriptProfilingEnabled):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::setJavaScriptProfilingEnabled):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setJavaScriptProfilingEnabled):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-06-02 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
NRWT needs a way to log to a file without including backspaces without needing --verbose
https://bugs.webkit.org/show_bug.cgi?id=60328
This change modifies NRWT so that if it can tell if it is
writing to a terminal, it will overwrite any messages printed
in the 'one-line-progress' and 'updates' print options, and if
not (or if --verbose was specified) the messages will be printed
without backspaces.
If the overwriting updates are disabled, one-line-progress will
only print every 10 seconds instead of on every update. This
ensures that we still get updates once in a while while the
tests are running, even if we're not in --verbose mode. Note
that one-line-progress will now still be printed in --verbose,
since it's useful as long as you're not flooded with the
updates.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2011-06-02 Martin Robinson <mrobinson@igalia.com>
Reviewed by Gustavo Noronha Silva.
[GTK] autogen.sh is run twice for each buld on the bots
https://bugs.webkit.org/show_bug.cgi?id=61951
* Scripts/webkitdirs.pm: Only check the previous autotools arguments
for the WebKit project. This prevents build-jsc runs from forcing subsequent
build-webkit runs to re-rerun autogen.sh.
2011-06-02 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r87946.
http://trac.webkit.org/changeset/87946
https://bugs.webkit.org/show_bug.cgi?id=61965
Cause NRWT to break (Requested by abarth|gardener on #webkit).
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-06-02 Dimitri Glazkov <dglazkov@chromium.org>
Reviewed by Darin Adler.
Add build logistics and plumbing for window.internals object.
https://bugs.webkit.org/show_bug.cgi?id=60313
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::bindJSObjectsToWindow): Added injection code.
* DumpRenderTree/mac/FrameLoadDelegate.mm:
(-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
2011-06-02 Hao Zheng <zhenghao@chromium.org>
Reviewed by Dirk Pranke.
Call cleanup() explicitly at the end of worker.run().
https://bugs.webkit.org/show_bug.cgi?id=61847
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-06-02 Dimitri Glazkov <dglazkov@chromium.org>
Unreviewed, rolling out r87926.
http://trac.webkit.org/changeset/87926
https://bugs.webkit.org/show_bug.cgi?id=60313
Fails to find WebCoreTestSupport.dylib on bots.
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::bindJSObjectsToWindow):
* DumpRenderTree/mac/FrameLoadDelegate.mm:
(-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
2011-05-21 Dimitri Glazkov <dglazkov@chromium.org>
Reviewed by Darin Adler.
Add build logistics and plumbing for window.internals object.
https://bugs.webkit.org/show_bug.cgi?id=60313
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::bindJSObjectsToWindow): Added injection code.
* DumpRenderTree/mac/FrameLoadDelegate.mm:
(-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
2011-06-02 Gyuyoung Kim <gyuyoung.kim@samsung.com>
Reviewed by Eric Seidel.
Adjust efl coding style to more efl directories
https://bugs.webkit.org/show_bug.cgi?id=61903
* Scripts/webkitpy/style/checker.py: Adjust efl coding style to efl directories.
2011-06-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
old-run-webkit-tests: add support for audio files
https://bugs.webkit.org/show_bug.cgi?id=57992
* Scripts/old-run-webkit-tests:
2011-06-01 Jochen Eisinger <jochen@chromium.org>
Reviewed by Adam Barth.
Implement layoutTestController.setImagesAllowed for controlling the respective behavior of the WebPermissionClient
https://bugs.webkit.org/show_bug.cgi?id=34314
* DumpRenderTree/DumpRenderTree.gypi:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setImagesAllowed):
(LayoutTestController::setStorageAllowed):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/chromium/WebPermissions.cpp: Copied from Tools/DumpRenderTree/chromium/WebPermissions.h.
(WebPermissions::WebPermissions):
(WebPermissions::~WebPermissions):
(WebPermissions::allowImages):
(WebPermissions::allowStorage):
(WebPermissions::setImagesAllowed):
(WebPermissions::setStorageAllowed):
(WebPermissions::reset):
* DumpRenderTree/chromium/WebPermissions.h:
2011-06-01 Noel Gordon <noel.gordon@gmail.com>
Reviewed by Tony Chang.
Test plugin should support event logging on the windows port.
https://bugs.webkit.org/show_bug.cgi?id=61721
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(handleEventWin): Well, handle them.
2011-06-01 Daniel Cheng <dcheng@chromium.org>
Reviewed by Tony Chang.
[chromium] Use correct file name for dragging out images.
https://bugs.webkit.org/show_bug.cgi?id=24887
* DumpRenderTree/chromium/EventSender.cpp:
(EventSender::EventSender):
(EventSender::dumpFilenameBeingDragged):
* DumpRenderTree/chromium/EventSender.h:
2011-06-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
nrwt: make webaudio work w/o content-length header
https://bugs.webkit.org/show_bug.cgi?id=61819
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/port/mock_drt.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-06-01 Viatcheslav Ostapenko <ostapenko.viatcheslav@nokia.com>
Reviewed by Andreas Kling.
[Qt][WK2] Add QGLWidget viewport support to MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=61802
Add command line parameter and menu item to MiniBrowser application
enabling use of QGLWidget for browser viewport.
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):
(BrowserWindow::toggleGLViewport):
* MiniBrowser/qt/BrowserWindow.h:
* MiniBrowser/qt/MiniBrowserApplication.cpp:
(MiniBrowserApplication::handleUserOptions):
* MiniBrowser/qt/MiniBrowserApplication.h:
(WindowOptions::WindowOptions):
2011-06-01 Martin Robinson <mrobinson@igalia.com>
Reviewed by Xan Lopez.
Print out autogen arguments when they differ from the last set, to help
with debugging build issues.
* Scripts/webkitdirs.pm: Print out previous and current build arguments
when they differ.
2011-06-01 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] run-launcher should run MiniBrowser in WebKit2
https://bugs.webkit.org/show_bug.cgi?id=61858
* Scripts/run-launcher: Check whether -2 command line option has
been passed and run MiniBrowser instead of GtkLauncher.
2011-05-31 Hao Zheng <zhenghao@chromium.org>
Reviewed by Dirk Pranke.
Fix exception in verbose printing mode.
https://bugs.webkit.org/show_bug.cgi?id=61723
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
2011-05-31 Yong Li <yoli@rim.com>
Reviewed by Eric Seidel.
https://bugs.webkit.org/show_bug.cgi?id=54807
We have been assuming plain bitfields (like "int a : 31") are always signed integers.
However some compilers can treat them as unsigned. For example, RVCT 4.0 states plain
bitfields (declared without either signed or unsigned qualifiers) are treats as unsigned.
http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0348c/Babjddhe.html
Although we can use "--signed-bitfields" flag to make RVCT 4.0 behave as most other compilers,
always using "signed"/"unsigned" qualifier to declare integral type bitfields is still a good
rule we should have in order to make our code independent from compilers and compiler flags.
* Scripts/webkitpy/style/checkers/cpp.py:
2011-05-31 Hironori Bono <hbono@chromium.org>
Reviewed by Eric Seidel.
Add a JavaScript function layoutTestController.setTextDirection().
https://bugs.webkit.org/show_bug.cgi?id=50952
This change adds a new JavaScript function setTextDirection() to the
LayoutTestController class of Safari and Chrome to write layout tests
that changes the text direction of an editable element.
* DumpRenderTree/LayoutTestController.cpp:
(setTextDirectionCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setTextDirection):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::setTextDirection):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::setTextDirection):
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::setTextDirection):
2011-05-31 Ademar de Souza Reis Jr <ademar.reis@openbossa.org>
Reviewed by Simon Hausmann.
[Qt] User-agent list of Qt test browsers is very outdated
https://bugs.webkit.org/show_bug.cgi?id=61531
The user can edit the user-agent manually, so there's no need to have
a super complete and up to date list, but it's nice to have it
minimally updated.
* MiniBrowser/qt/useragentlist.txt:
* QtTestBrowser/useragentlist.txt:
2011-03-30 Martin Robinson <mrobinson@igalia.com>
Reviewed by Adam Roben.
[GTK] [WebKit2] Implement a basic WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=57068
Add an implementation of WebKitTestRunner for GTK+.
* Scripts/build-webkittestrunner: Added knowledge of GTK+ TestRunner.
* Scripts/old-run-webkit-tests: Ditto.
* Scripts/run-launcher: Ditto.
* Scripts/webkitdirs.pm: Ditto.
* WebKitTestRunner/GNUmakefile.am: Added.
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
* WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp: Added.
* WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
* WebKitTestRunner/InjectedBundle/gtk/LayoutTestControllerGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
* WebKitTestRunner/PlatformWebView.h:
* WebKitTestRunner/gtk/PlatformWebViewGtk.cpp: Added.
* WebKitTestRunner/gtk/TestControllerGtk.cpp: Added.
* WebKitTestRunner/gtk/TestInvocationGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
(WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
* WebKitTestRunner/gtk/main.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
(main): Added.
2011-05-31 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: doesn't wait for children if it gets a ctrl-c
https://bugs.webkit.org/show_bug.cgi?id=60241
* Scripts/new-run-webkit-tests:
2011-05-31 Eric Seidel <eric@webkit.org>
Reviewed by David Kilzer.
webkitpy should create zips with zip -9
https://bugs.webkit.org/show_bug.cgi?id=61789
Dave Kilzer suggested we add this to make zips uploaded to bugzilla smaller.
* Scripts/webkitpy/common/system/workspace.py:
* Scripts/webkitpy/common/system/workspace_unittest.py:
2011-05-31 Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[Qt] Implement find feature for QtTestBrowser
https://bugs.webkit.org/show_bug.cgi?id=61425
Buildfix for --minimal and CONFIG+=qt_minimal build.
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::LauncherWindow):
(LauncherWindow::createChrome):
* QtTestBrowser/launcherwindow.h:
2011-05-31 Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[Qt] Implement find feature for QtTestBrowser
https://bugs.webkit.org/show_bug.cgi?id=61425
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::LauncherWindow):
(LauncherWindow::createChrome):
(LauncherWindow::showFindBar):
(LauncherWindow::find):
* QtTestBrowser/launcherwindow.h:
2011-05-30 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Implement JavaScript dialogs in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=61725
* MiniBrowser/gtk/BrowserWindow.c:
(WKStringGetCString): Helper function to get a UTF-8 c string from
a WKStringRef.
(WKURLGetCString): Helper function to get a UTF-8 c string from a
WKURLRef.
(browserWindowUpdateURL): Update to use WKURLGetCString.
(didReceiveTitleForFrame): Update to use WKStringGetCString.
(createMessageDialog): Create a GtkMessageDialog.
(runJavaScriptAlert):
(runJavaScriptConfirm):
(runJavaScriptPrompt):
(browserWindowUIClientInit):
2011-05-29 Yuta Kitamura <yutak@chromium.org>
Reviewed by Kent Tamura.
WebSocket closing handshake
https://bugs.webkit.org/show_bug.cgi?id=35721
* Scripts/webkitpy/thirdparty/__init__.py:
Pull in pywebsocket 0.6b1. We need to update pywebsocket
to get the right behavior of closing handshake.
2011-05-29 Daniel Bates <dbates@rim.com>
Reviewed by David Kilzer.
REGRESSION (r86515): svn-apply ignores diffs that omit line count in chunk range
https://bugs.webkit.org/show_bug.cgi?id=61162
Fixes an issue where svn-apply may ignore a diff that contains a chunk range line
that omits a line count. In particular, the chunk range regular expression does
not match a chunk range line that omits a line count. GNU diff(1) will omit the
line count in the chunk range if the line count is exactly 1. For example, appending
a new line to the end of an existing file F that contains exactly one line of text will
be represented in a diff with a chunk range line that omits the line count for F.
* Scripts/VCSUtils.pm:
(parseChunkRange): Added.
* Scripts/webkitperl/VCSUtils_unittest/parseChunkRange.pl: Added.
* Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
- Added unit test "Git: Append new line to the end of an existing file".
2011-05-28 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
EWS builds patches that fail to build twice, which seems useless and slows down the bots
https://bugs.webkit.org/show_bug.cgi?id=55585
This patch switches all the early warning system bots over to the new
PatchAnalysisTask-based infrastructure. This patch makes these bots
more efficient (in the case where patches fail to build) and paves the
way for running tests on these bots!
* Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
* Scripts/webkitpy/tool/bot/earlywarningsystemtask.py:
* Scripts/webkitpy/tool/bot/patchanalysistask.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
* Scripts/webkitpy/tool/commands/queues.py:
2011-05-28 Kenichi Ishibashi <bashi@chromium.org>
Reviewed by Kent Tamura.
Add bashi as a committer.
https://bugs.webkit.org/show_bug.cgi?id=61685
* Scripts/webkitpy/common/config/committers.py:
2011-05-27 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
When checking whether the tree is red, the EWS posts a link to the wrong log
https://bugs.webkit.org/show_bug.cgi?id=61072
We need to cache the original script error because that contains the
failure log we want to upload. If we don't cache that script error,
self._script_error will get overwritten when we sanity check the clean
tree (and it also has test failures).
* Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
- This testing approach is slightly goofy. We'd like to use
assertRaisesRegexp, but that's not available until Python 2.7.
* Scripts/webkitpy/tool/bot/patchanalysistask.py:
2011-05-27 Jochen Eisinger <jochen@chromium.org>
Reviewed by Adam Barth.
Add layoutTestController.setStorageAllowed() to control whether access
to the localStorage API is enabled via the WebPermissionClient
https://bugs.webkit.org/show_bug.cgi?id=61581
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setStorageAllowed):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
(TestShell::createNewWindow):
* DumpRenderTree/chromium/TestShell.h:
* DumpRenderTree/chromium/WebPermissions.h: Added.
(WebPermissions::WebPermissions):
(WebPermissions::allowStorage):
(WebPermissions::setStorageAllowed):
(WebPermissions::reset):
2011-05-27 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
NRWT: clean up metered_stream code in preparation for 'nooverwriting' patch
https://bugs.webkit.org/show_bug.cgi?id=60326
This patch removes a lot of the complexity from the
metered_stream implementation that was unnecessary since there
was only one caller and the logic could be coordinated better.
There should be no functional changes in this patch, just code
getting deleted and cleaned up.
* Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
* Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2011-05-27 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
NRWT: minor cleanup in printing module
https://bugs.webkit.org/show_bug.cgi?id=60329
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
2011-05-27 Steve Lacey <sjl@chromium.org>
Reviewed by David Levin.
Add sjl@chromium.org as a committer.
https://bugs.webkit.org/show_bug.cgi?id=61672
* Scripts/webkitpy/common/config/committers.py:
2011-05-27 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
NRWT: remove --print detailed-progress
https://bugs.webkit.org/show_bug.cgi?id=60324
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-05-27 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
NRWT: debug messages from the workers are being logged twice
https://bugs.webkit.org/show_bug.cgi?id=60428
It looks like when the workers are run in separate processes
we end up getting two copies of every log message they print.
This has to do with the multiprocessing module on UNIX cloning
the log configuration in a way I wasn't expecting, and so two
log handlers end up getting registered.
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-05-27 Adam Roben <aroben@apple.com>
Always decode tester names from location.hash
When navigating directly to a tester-specific URL (e.g.,
<http://build.webkit.org/TestFailures/#/Windows%207%20Release%20(Tests)>), you could end up
with a URL-encoded tester name in the page, and then the encoded name could make its way
into Bugzilla bugs. Bad!
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype.parseHash): Pass the tester name we got from location.hash through
decodeURIComponent before trying to use it.
2011-05-27 Adam Roben <aroben@apple.com>
Remove testing code that snuck into r87528
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForNewBugLink): Remove a line I was using for some testing.
2011-05-27 Adam Roben <aroben@apple.com>
Add links to TestFailures to easily file bugs about failing tests
Fixes <http://webkit.org/b/61636> <rdar://problem/9514886> TestFailures page should have
links to file bugs for failing tests
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.resultsPageURL): New helper function to return the URL to results.html
for a particular build.
(Builder.prototype._getFailingTests): Use resultsPageURL.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayBuilder): Add links to file bugs about each set of
failures.
(ViewController.prototype._domForBuildName): Use resultsPageURL.
(ViewController.prototype._domForNewBugLink): Added. Creates an element that represents the
new bug link. The bug has a title, description, and URL already filled in for you!
2011-05-27 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r87464.
http://trac.webkit.org/changeset/87464
https://bugs.webkit.org/show_bug.cgi?id=61643
client-close/server-close tests are flaky on Chromium Mac and
Windows (Requested by aklein on #webkit).
* Scripts/webkitpy/thirdparty/__init__.py:
2011-05-27 Alexey Proskuryakov <ap@apple.com>
Added Douglas Davidson to contributors list for easier CC'ing.
* Scripts/webkitpy/common/config/committers.py:
2011-05-26 Mihai Parparita <mihaip@chromium.org>
Reviewed by Adam Barth.
Fix worldID and destinationDomain argument names
https://bugs.webkit.org/show_bug.cgi?id=61571
As part of working on r87423, I noticed a couple of inconsistencies in
argument names:
- We would use worldId in .h files but worldID in .cpp files.
Standardize on the latter
- SecurityOrigin::addOriginAccessWhitelistEntry would take a
destinationDomains argument, even though the actual parameter was for
a single domain (renamed to be singular).
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::evaluateScriptInIsolatedWorld):
2011-05-26 Yuta Kitamura <yutak@chromium.org>
Reviewed by Kent Tamura.
WebSocket closing handshake
https://bugs.webkit.org/show_bug.cgi?id=35721
* Scripts/webkitpy/thirdparty/__init__.py:
Pull in pywebsocket 0.6b1. We need to update pywebsocket
to get the right behavior of closing handshake.
2011-05-26 Qi Zhang <qi.2.zhang@nokia.com>
Reviewed by Andreas Kling.
WebKitTestRunner needs layoutTestController.setAuthorAndUserStylesEnabled
https://bugs.webkit.org/show_bug.cgi?id=42676
Implemented layoutTestController.setAuthorAndUserStylesEnabled for WebKitTestRunner.
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setAuthorAndUserStylesEnabled):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-05-26 Ademar de Souza Reis Jr. <ademar.reis@openbossa.org>
Reviewed by Andreas Kling.
[Qt] QtTestBrowser "Switch User Agent" dialog is broken
https://bugs.webkit.org/show_bug.cgi?id=61527
The setMaximumHeight() call is unecessary and breaks the dialog at
least on my setup.
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::showUserAgentDialog):
2011-05-02 Robert Hogan <robert@webkit.org>
Reviewed by Adam Roben.
[Qt] Allow popup windows from plugins when initiated by a user gesture
https://bugs.webkit.org/show_bug.cgi?id=41292
Support parameters introduced in plugin-initiate-popup-window.html
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_new_instance):
(webkit_test_plugin_handle_event):
2011-05-26 Mihai Parparita <mihaip@chromium.org>
Reviewed by Adam Barth.
Support cross-origin XMLHttpRequest in isolated worlds
https://bugs.webkit.org/show_bug.cgi?id=59843
Expose setIsolatedWorldSecurityOrigin in LayoutTestController.
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setIsolatedWorldSecurityOrigin):
* DumpRenderTree/chromium/LayoutTestController.h:
2011-05-25 Brian Weinstein <bweinstein@apple.com>
Reviewed by Adam Roben.
WebKit2: Status bar, toolbar, and menu bar checks should be in the injected bundle
https://bugs.webkit.org/show_bug.cgi?id=61474
<rdar://problem/9468337>
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage): Add empty entries in the WKBundlePageUIClient.
2011-05-26 Andreas Kling <kling@webkit.org>
Unreviewed, correct address to webkit-committers mailing list.
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
2011-05-25 Adam Roben <aroben@apple.com>
Don't assume that builds finish in the order they were started
Fixes <http://webkit.org/b/61464> TestFailures page shows failures from a not-most-recent
build when one slave is taking a long time to finish a build
Reviewed by Antti Koivisto.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getMostRecentCompletedBuildNumber): Don't assume that all builds that are
newer than the oldest in-progress build are also themselves in progress. Just check whether
each build is actually in progress (as signified by its presence in the currentBuilds set).
2011-05-25 Adam Roben <aroben@apple.com>
Correctly detect cases where only a single test failed/crashed/timed out
Fixes <http://webkit.org/b/61463> TestFailures page doesn't show testers with only a single
failing test
Reviewed by Darin Adler.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getNumberOfFailingTests): Don't require "test cases" to be plural. It
will be singular when only a single test fails/crashes/times out.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayTesters): Only pluralize "test" when more than one test
failed/crashed/timed out.
2011-05-26 Tom Hudson <tomhudson@google.com>
Reviewed by Tony Chang.
Add flags to Chromium ImageDiff to write image comparison metrics on
stdout.
https://bugs.webkit.org/show_bug.cgi?id=60569
* DumpRenderTree/chromium/ImageDiff.cpp:
(Image::createFromStdin): Fix spelling.
(maxOf3): New function to do 3-way maximum.
(getRedComponent): Extract red component from ABGR packed encoding.
(getGreenComponent): Extract green component from ABGR packed encoding.
(getBlueComponent): Extract blue component from ABGR packed encoding.
(weightedPercentageDifferent): Determine % of pixels different in two
images multiplied by intensity difference of each pixel.
(printHelp): Document new arguments.
(compareImages): Parameterize.
(untestedCompareImages): Parameterize.
(diffImages): Parameterize; write results on stdout if requested.
(main): Parse new arguments.
2011-05-25 Qi Zhang <qi.2.zhang@nokia.com>
Reviewed by Simon Hausmann.
WebKitTestRunner needs layoutTestController.setPrivateBrowsingEnabled
https://bugs.webkit.org/show_bug.cgi?id=42697
Implemented layoutTestController.setPrivateBrowsingEnabled for WebKitTestRunner.
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setPrivateBrowsingEnabled):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-05-25 Jon Honeycutt <jhoneycutt@apple.com>
REGRESSION (WebKit2): Crash in Flash on USA Today photo gallery
https://bugs.webkit.org/show_bug.cgi?id=61428
<rdar://problem/9457006>
Reviewed by Adam Roben.
The crash occurs when Flash posts a message to a window that it
creates, and in processing the message, it calls NPN_Evaluate to
evaluate JavaScript that removes the plug-in from the page. Flash then
crashes when we return to Flash code.
This test emulates that behavior.
* DumpRenderTree/TestNetscapePlugIn/win/CallJSThatDestroysPlugin.cpp: Added.
(CallJSThatDestroysPlugin::CallJSThatDestroysPlugin):
Initialize member vars.
(CallJSThatDestroysPlugin::~CallJSThatDestroysPlugin):
Remove our custom property from the message window, and destroy it.
(CallJSThatDestroysPlugin::NPP_Destroy):
Set m_isDestroyed, log that the plug-in was destroyed, and notify the
layout test controller that we're done.
(wndProc):
Get the PluginTest object, and call its runTest() function.
(CallJSThatDestroysPlugin::NPP_New):
Setup the test: register a class for the message-only window, create
it, and post a message to it to run the test.
(CallJSThatDestroysPlugin::runTest):
Execute JS that removes the plug-in from the page, and if we're not
destroyed, log a success message.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
Add new test to project.
2011-05-25 Tony Chang <tony@chromium.org>
Reviewed by Adam Barth.
webkitpy's NetworkTransaction shouldn't use mechanize
https://bugs.webkit.org/show_bug.cgi?id=61450
mechanize.HTTPError is just an alias for urllib2.HTTPError, so
use urllib2 directly. Re-add NetworkTransaction to
test_results_uploader.py (reverting r87124).
* Scripts/webkitpy/common/net/networktransaction.py:
* Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
2011-05-25 Kevin Ollivier <kevino@theolliviers.com>
[Chromium] Unreviewed build fix. Do not define WTF/JS_EXPORT_PRIVATE to JS_EXPORTDATA
until the port begins using those export macros.
* DumpRenderTree/chromium/config.h:
2011-05-25 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r87257.
http://trac.webkit.org/changeset/87257
https://bugs.webkit.org/show_bug.cgi?id=61457
This broke layout tests, see bug 61431. (Requested by
hwennborg on #webkit).
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::clearAllDatabases):
* DumpRenderTree/chromium/LayoutTestController.h:
2011-05-25 Qi Zhang <qi.2.zhang@nokia.com>
Reviewed by Andreas Kling.
[Qt] fast/css/disabled-author-styles.html failed
https://bugs.webkit.org/show_bug.cgi?id=61438
Implemented API setAuthorAndUserStylesEnabled in LayoutTestControllerQt.
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::resetSettings):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setAuthorAndUserStylesEnabled):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-05-25 Adam Roben <aroben@apple.com>
Identify, rather than skip, builds where run-webkit-tests exited early due to too many failures
Fixes <http://webkit.org/b/61441> TestFailures page should show when run-webkit-tests
started exiting early due to too many crashes, timeouts, or failures
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getNumberOfFailingTests): Rather than returning -1 ("unknown") when
run-webkit-tests exits early due to too many failures, pass that information in another
argument to the callback.
(Builder.prototype.startFetchingBuildHistory): Updated the documentation comment to reflect
the new structure of the object passed to the callback.
(Builder.prototype._getFailingTests): Updated to expect a tooManyFailures boolean from
getNumberOfFailingTests and to pass that along to our own callbacks.
(Builder.prototype._incorporateBuildHistory): Updated to expect a tooManyFailures boolean
from _getFailingTests and to store that value in the history object.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(.info): Added.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayBuilder): Updated for change in structure to the history
object and to add a note when run-webkit-tests exited early due to too many failures.
(ViewController.prototype._displayTesters): Renamed testersAndFailureCounts to
latestBuildInfos since each entry now contains more than just the tester and failure count.
Now displays a message for testers where the latest build exited early due to too many
failures. Updated to expect a tooManyFailures boolean from getNumberOfFailingTests and to
store that value in latestBuildInfos.
2011-05-24 Keishi Hattori <keishi@webkit.org>
Reviewed by Kent Tamura.
Disable textfield implementation of <input type=color>. Add INPUT_COLOR feature flag. Add input color sanitizer.
https://bugs.webkit.org/show_bug.cgi?id=61273
* Scripts/build-webkit: Added INPUT_COLOR feature flag.
2011-05-24 Greg Simon <gregsimon@chromium.org>
Reviewed by Dimitri Glazkov.
Control Indexeddb backends from LayoutTestController
https://bugs.webkit.org/show_bug.cgi?id=61000
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setOverrideIndexedDBBackingStore):
(LayoutTestController::clearAllDatabases):
* DumpRenderTree/chromium/LayoutTestController.h:
2011-05-24 Robert Hogan <robert@webkit.org>
Reviewed by Ryosuke Niwa.
[Qt] Fix fast/events/selectstart-by-single-click-with-shift.html
https://bugs.webkit.org/show_bug.cgi?id=61393
* DumpRenderTree/qt/EventSenderQt.h:
* DumpRenderTree/qt/EventSenderQt.cpp:
(getModifiers): static function for interpreting modifiers
(EventSender::mouseDown): Respect keyboard modifiers
(EventSender::keyDown): use new static function
(EventSender::sendOrQueueEvent): whitespace fix
2011-05-24 Jay Civelli <jcivelli@chromium.org>
Reviewed by Adam Barth.
Adding MHTML reading support.
https://bugs.webkit.org/show_bug.cgi?id=7168
* Scripts/build-webkit:
* Scripts/old-run-webkit-tests:
* Scripts/webkitperl/features.pm:
* Scripts/webkitpy/layout_tests/port/test_files.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-05-24 Stephen White <senorblanco@chromium.org>
Reviewed by Kenneth Russell.
Remove most style checks on Source/ThirdParty/glu, since it doesn't
follow WebKit style.
https://bugs.webkit.org/show_bug.cgi?id=61365
* Scripts/webkitpy/style/checker.py:
2011-05-24 Adam Roben <aroben@apple.com>
Don't allow forcing builds through the build.webkit.org web interface
Someone has been forcing bogus builds.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-05-24 Qi Zhang <qi.2.zhang@nokia.com>
Reviewed by Adam Roben.
WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
https://bugs.webkit.org/show_bug.cgi?id=42669
Fixes the regression issue caused by patch 94265. The patch didn't intialize the setting before the
layouttest starting, it will make the test cases after editing/execCommand/clipboard-access.html failed.
Because that case changed the setting.
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
2011-05-24 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[ORWT]Make --skipped=only option work with parameters with LayoutTests/ prefix
https://bugs.webkit.org/show_bug.cgi?id=61052
* Scripts/old-run-webkit-tests:
2011-05-24 Adam Roben <aroben@apple.com>
Make TestFailures show how many tests are failing on each tester, and omit testers with no failures
Fixes <http://webkit.org/b/61063> <rdar://problem/9460533> TestFailures page shows testers
that don't have any failing tests, which isn't useful
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
(Buildbot.prototype.getTesters): Renamed from getTesterNames. Now returns Builder objects
instead of name strings.
(Buildbot.prototype._buildersForNames): Added. Helper function to convert an array of
builder names into an array of builders.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getMostRecentCompletedBuildNumber): Added. Returns the build number of
the most recently completed build, or -1 if there is no such build.
(Builder.prototype.getNumberOfFailingTests): Added. Returns the number of tests that failed
in the given build, or -1 if the number could not be determined. Some of this code came from
_getFailingTests.
(Builder.prototype._getBuildJSON): Added. Code came from _getFailingTests.
(Builder.prototype._getFailingTests): Changed to use new _getBuildJSON and
getNumberOfFailingTests functions.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayTesters): Get the current number of test failures for each
tester and show it in the list. Omit testers that have no failures at all. We keep the
testers and failure counts in an array and sort it before displaying the current data, as
the order in which data will be fetched is unpredictable.
2011-05-23 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[nrwt] don't use NetworkTranslation when uploading results
https://bugs.webkit.org/show_bug.cgi?id=61315
This avoids autoinstalling mechanize. It didn't matter anyway
because test_results_uploader.py wasn't using mechanize to make the
request so NetworkTransaction::run wasn't doing anything extra
(it only matters if it catches mechanize.HTTPError).
* Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
2011-05-23 Tony Chang <tony@chromium.org>
Reviewed by Dirk Pranke.
[chromium] fix the rebaseline tool to work with Linux GPU failures
https://bugs.webkit.org/show_bug.cgi?id=61312
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-05-23 Darin Adler <darin@apple.com>
Ignore generated .pyc files in a directory full of python code.
* Scripts/webkitpy/common/checkout/scm: Added property svn:ignore.
2011-05-23 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
sheriffbot should give more details about the failures in IRC
https://bugs.webkit.org/show_bug.cgi?id=61233
With this patch, sheriffbot will annonce the set of failing tests,
which might help folks triage the problem.
* Scripts/webkitpy/tool/commands/sheriffbot.py:
* Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
2011-05-23 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r87078.
http://trac.webkit.org/changeset/87078
https://bugs.webkit.org/show_bug.cgi?id=61294
Patch contains errors and seems likely to break a bunch of
bots (Requested by abarth on #webkit).
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2011-05-23 Alice Boxhall <aboxhall@chromium.org>
Reviewed by Ojan Vafai.
Convert json_results_generator.py to output version 4 JSON.
https://bugs.webkit.org/show_bug.cgi?id=60869
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2011-05-23 Patrick Gansterer <paroga@webkit.org>
Reviewed by Adam Barth.
Move generation of shell_command from port to executive
https://bugs.webkit.org/show_bug.cgi?id=61251
After move it's possible to use the code in other places too.
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/common/system/executive.py:
2011-05-23 Qi Zhang <qi.2.zhang@nokia.com>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
https://bugs.webkit.org/show_bug.cgi?id=42669
Implement setJavaScriptCanAccessClipboard in WebKitTestRunner.
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setJavaScriptCanAccessClipboard):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-05-22 Robert Hogan <robert@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
Fix policyDelegate in Qt DRT
https://bugs.webkit.org/show_bug.cgi?id=61247
Use the 'policy delegate' implemented in FrameLoaderClient
for layout tests. The partial implementation in DumpRenderTreeQt
is redundant, so remove it.
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::acceptNavigationRequest):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::waitForPolicyDelegate):
2011-05-22 Patrick Gansterer <paroga@webkit.org>
Reviewed by Darin Adler.
Use double quotes for filename arguments
https://bugs.webkit.org/show_bug.cgi?id=61250
Windows does not support single quotes on the command line,
so replace them with double quotes.
* Scripts/prepare-ChangeLog:
2011-05-20 Simon Fraser <simon.fraser@apple.com>
Reviewed by Sam Weinig.
WebKitTestRunner needs layoutTestController.pauseTransitionAtTimeOnElementWithId
https://bugs.webkit.org/show_bug.cgi?id=42550
Implement pauseTransitionAtTimeOnElementWithId in WebKitTestRunner.
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
Copy the code for pausing animations, FIXME comments and all.
(WTR::LayoutTestController::pauseTransitionAtTimeOnElementWithId):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-05-20 Alok Priyadarshi <alokp@chromium.org>
Reviewed by Stephen White.
Adding myself to the committers list.
* Scripts/webkitpy/common/config/committers.py:
2011-05-19 Adam Roben <aroben@apple.com>
Test that the WebKit2 UI process doesn't crash when starting a download
Test for <http://webkit.org/b/61142> <rdar://problem/9471680> REGRESSION (r86812): Crash
(preceded by assertion) in fastMalloc when downloading a file
Reviewed by Darin Adler.
* TestWebKitAPI/Tests/WebKit2/18-characters.html: Added.
* TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp: Added.
(TestWebKitAPI::decidePolicyForNavigationAction): Start a download.
(TestWebKitAPI::decideDestinationWithSuggestedFilename): Record that the download was
started, cancel the download, and return a bogus string.
(TestWebKitAPI::setContextDownloadClient):
(TestWebKitAPI::setPagePolicyClient):
Simple helper functions.
(TestWebKitAPI::TEST): Load 18-characters.html, which should trigger a download thanks to
our policy client, and run until we know that the download was started. If we haven't
crashed, we win!
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
* TestWebKitAPI/win/copy-resources.cmd:
Added new files.
2011-05-20 Kent Tamura <tkent@chromium.org>
Reviewed by Ryosuke Niwa.
Fix style errors in DumpRenderTree/chromium/.
https://bugs.webkit.org/show_bug.cgi?id=61172
* DumpRenderTree/chromium/CppBoundClass.cpp:
(CppBoundClass::getAsCppVariant):
(CppBoundClass::bindToJavascript):
* DumpRenderTree/chromium/CppBoundClass.h:
(CppBoundClass::GetterCallback::~GetterCallback):
(CppBoundClass::CppBoundClass):
(CppBoundClass::Callback::~Callback):
(CppBoundClass::MemberCallback::MemberCallback):
(CppBoundClass::MemberCallback::~MemberCallback):
(CppBoundClass::MemberGetterCallback::MemberGetterCallback):
(CppBoundClass::MemberGetterCallback::~MemberGetterCallback):
* DumpRenderTree/chromium/CppVariant.h:
* DumpRenderTree/chromium/DRTDevToolsAgent.h:
(DRTDevToolsAgent::~DRTDevToolsAgent):
* DumpRenderTree/chromium/DRTDevToolsClient.cpp:
(DRTDevToolsClient::sendFrontendLoaded):
* DumpRenderTree/chromium/DRTDevToolsClient.h:
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/EventSender.cpp:
(SavedEvent::SavedEvent):
(EventSender::EventSender):
(EventSender::keyDown):
* DumpRenderTree/chromium/EventSender.h:
* DumpRenderTree/chromium/ImageDiff.cpp:
(Image::Image):
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::WorkQueue::reset):
(WorkItemBackForward::WorkItemBackForward):
(WorkItemLoadingScript::WorkItemLoadingScript):
(WorkItemNonLoadingScript::WorkItemNonLoadingScript):
(WorkItemLoad::WorkItemLoad):
(WorkItemLoadHTMLString::WorkItemLoadHTMLString):
(LayoutTestController::pathToLocalResource):
* DumpRenderTree/chromium/LayoutTestController.h:
(LayoutTestController::WorkItem::~WorkItem):
(LayoutTestController::WorkQueue::WorkQueue):
* DumpRenderTree/chromium/MockSpellCheck.cpp:
(MockSpellCheck::MockSpellCheck):
(MockSpellCheck::~MockSpellCheck):
* DumpRenderTree/chromium/MockSpellCheck.h:
* DumpRenderTree/chromium/NotificationPresenter.h:
(NotificationPresenter::NotificationPresenter):
* DumpRenderTree/chromium/Task.cpp:
(WebTask::WebTask):
* DumpRenderTree/chromium/Task.h:
(TaskList::TaskList):
* DumpRenderTree/chromium/TestEventPrinter.cpp:
(DRTPrinter::DRTPrinter):
(TestShellPrinter::TestShellPrinter):
* DumpRenderTree/chromium/TestNavigationController.cpp:
(TestNavigationEntry::TestNavigationEntry):
(TestNavigationEntry::~TestNavigationEntry):
(TestNavigationController::TestNavigationController):
(TestNavigationController::reload):
(TestNavigationController::loadEntry):
(TestNavigationController::didNavigateToEntry):
* DumpRenderTree/chromium/TestNavigationController.h:
(TestShellExtraData::TestShellExtraData):
* DumpRenderTree/chromium/TestShell.cpp:
(dumpHistoryItem):
* DumpRenderTree/chromium/TestShell.h:
(TestParams::TestParams):
* DumpRenderTree/chromium/TestShellWin.cpp:
(TestShell::waitTestFinished):
* DumpRenderTree/chromium/TestWebWorker.h:
(TestWebWorker::startWorkerContext):
(TestWebWorker::terminateWorkerContext):
(TestWebWorker::postMessageToWorkerContext):
(TestWebWorker::clientDestroyed):
(TestWebWorker::postMessageToWorkerObject):
(TestWebWorker::postExceptionToWorkerObject):
(TestWebWorker::postConsoleMessageToWorkerObject):
(TestWebWorker::confirmMessageFromWorkerObject):
(TestWebWorker::reportPendingActivity):
(TestWebWorker::workerContextClosed):
(TestWebWorker::~TestWebWorker):
* DumpRenderTree/chromium/TextInputController.cpp:
* DumpRenderTree/chromium/WebThemeControlDRTWin.h:
* DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
* DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
(WebThemeEngineDRTWin::paintButton):
(WebThemeEngineDRTWin::paintMenuList):
(WebThemeEngineDRTWin::paintTrackbar):
* DumpRenderTree/chromium/WebThemeEngineDRTWin.h:
(WebThemeEngineDRTWin::WebThemeEngineDRTWin):
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::navigate):
(WebViewHost::updateSessionHistory):
(WebViewHost::paintInvalidatedRegion):
* DumpRenderTree/chromium/WebViewHost.h:
2011-05-19 Dmitry Lomov <dslomov@google.com>
Reviewed by Adam Roben.
Detect hangs in run-api-tests
https://bugs.webkit.org/show_bug.cgi?id=48043
* Scripts/run-api-tests: Added test timeouts
2011-05-19 Dmitry Lomov <dslomov@google.com>
Reviewed by Adam Roben.
run-api-tests should run one test per process
https://bugs.webkit.org/show_bug.cgi?id=61088
* Scripts/run-api-tests: Resurrecting the previous revison of this file, with fixes to system call under Windows,
return code, and parsing GTest output format.
2011-05-15 Robert Hogan <robert@webkit.org>
Reviewed by Antonio Gomes.
[Qt][GTK] plugins/get-url-with-javascript-url.html fails
https://bugs.webkit.org/show_bug.cgi?id=60834
Fix unix test plugin for plugins/get-url-with-javascript-url.html
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_new_stream):
(webkit_test_plugin_write_ready):
(webkit_test_plugin_write):
2011-05-19 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r86869, r86873, r86875, and r86877.
http://trac.webkit.org/changeset/86869
http://trac.webkit.org/changeset/86873
http://trac.webkit.org/changeset/86875
http://trac.webkit.org/changeset/86877
https://bugs.webkit.org/show_bug.cgi?id=61139
broke builds and debug DRT (Requested by rniwa on #webkit).
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::bindJSObjectsToWindow):
* DumpRenderTree/mac/FrameLoadDelegate.mm:
(-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
2011-05-18 Dimitri Glazkov <dglazkov@chromium.org>
Reviewed by Darin Adler.
Add build logistics and plumbing for window.internals object.
https://bugs.webkit.org/show_bug.cgi?id=60313
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::bindJSObjectsToWindow): Added injection code.
* DumpRenderTree/mac/FrameLoadDelegate.mm:
(-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
2011-05-19 Zsolt Fehér <h490267@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[Qt] Implement eventSender.scalePageBy
https://bugs.webkit.org/show_bug.cgi?id=60015
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
* DumpRenderTree/qt/EventSenderQt.cpp:
(EventSender::scalePageBy):
* DumpRenderTree/qt/EventSenderQt.h:
2011-05-19 Kenichi Ishibashi <bashi@chromium.org>
Reviewed by Kent Tamura.
[Chromium] IME candidate window appears wrong position in an iframe
https://bugs.webkit.org/show_bug.cgi?id=61023
Call the focused frame's firstRectForCharacterRange() instead of the
main frame so that DRT behaves as the same as Chromium.
* DumpRenderTree/chromium/TextInputController.cpp:
(TextInputController::firstRectForCharacterRange):
2011-05-18 Adam Roben <aroben@apple.com>
Add a new page to build.webkit.org to help find when tests started failing
The page is accessible at <http://build.webkit.org/TestFailures/>. It is pretty minimalist
right now, but already shows some useful information. It's somewhat similar to webkit-patch
failure-reason and sheriffbot, and perhaps can be combined with them eventually. It's a
little more convenient than either of them, though, because it's all done in the browser
(and thus it's easy to go directly to the relevant test results).
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js: Added.
(Buildbot): This class represents a Buildbot server.
(Buildbot.prototype.buildURL): Returns the URL for the summary page for a particular build.
(Buildbot.prototype.builderNamed): Returns a Builder with the given name.
(Buildbot.prototype.getTesterNames): Fetches the names of all testers and passes them to the
callback.
(Buildbot.prototype.parseBuildName): Breaks up a build name into its constituent parts. Must
be implemented by a derived class that understands this server's build naming scheme.
(Buildbot.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
particular build.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js: Added.
(Builder): This class represents one builder on the buildbot.
(Builder.prototype.buildURL): Returns the URL for the summary page for a particular build.
(Builder.prototype.failureDiagnosisTextAndURL): Returns data that provides a little more
information about a particular test failure.
(Builder.prototype.startFetchingBuildHistory): Periodically calls the callback with
information about when tests started failing.
(Builder.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
particular build.
(Builder.prototype._getBuildNames): Fetches the names of all builds and passes them to the
callback.
(Builder.prototype._getFailingTests): Fetches the results.html page for the given build and
extracts all the failing tests listed in it, passing them to the callback.
(Builder.prototype._incorporateBuildHistory): Gets the failing tests for the specified
build, merges them into the build history, and calls the callback telling it whether the
next build should be fetched to provide more information.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
Added. Just some simple styles.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js: Added.
(createDefinitionList): Takes an array of pairs and turns them into a DL element.
(getResource): Wrapper around XMLHttpRequest.
(Array.prototype.findFirst): Finds the first element matching the given predicate and
returns it.
(Array.prototype.last): Returns the last element of the array.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js: Added.
(ViewController): This class contains the main logic for displaying the page.
(ViewController.loaded): Just calls through to parseHash.
(ViewController.parseHash): Either starts analyzing failures on a particular builder, or
shows the list of all testers so one can be chosen. This function is called when the page
loads and whenever we get a hashchange event.
(ViewController._displayBuilder): Asks the builder to fetch build history, and displays it
as it is fetched. The display ends up grouping tests by when they started failing.
(ViewController._displayTesters): Gets the list of testers and displays it.
(ViewController._domForBuildName):
(ViewController._domForFailedTest):
Helper functions to create descriptions and links for a particular build or failed test.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js: Added.
(WebKitBuildbot): Calls up to the base class constructor with the correct base URL.
(WebKitBuildbot.prototype.parseBuildName): Parses a build.webkit.org-style build name.
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Added. Just
loads all the files and sets up a ViewController, which does the rest.
* BuildSlaveSupport/build.webkit.org-config/templates/root.html: Added a link to the new
page.
2011-05-18 Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[Qt] Implement layoutTestController.setValueForUser()
https://bugs.webkit.org/show_bug.cgi?id=60956
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setValueForUser):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-05-16 Robert Hogan <robert@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
plugins/invalidate_rect.html fails on linux ports
- Make the unix test netscape plugin recognize the onPaintEvent
and windowedPlugin parameters.
https://bugs.webkit.org/show_bug.cgi?id=54051
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(handleEventX11):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::displayInvalidatedRegion):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_new_instance):
(webkit_test_plugin_handle_event):
2011-05-17 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[chromium] move Lucid 64 bit results into LayoutTests/platform/chromium-linux
https://bugs.webkit.org/show_bug.cgi?id=60895
Update the tools to handle the move. Lucid 32 results now go in
chromium-linux-x86 and the default platform on Linux is now x86_64.
* Scripts/webkitpy/layout_tests/deduplicate_tests.py: Default to x86_64
* Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Default to Lucid
* Scripts/webkitpy/layout_tests/port/base.py: ditto
* Scripts/webkitpy/layout_tests/port/chromium_linux.py: Update directory fallback and default to Lucid 64
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
* Scripts/webkitpy/layout_tests/port/test.py: Update tests to default to x86_64
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Update bot names.
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Whitespace cleanup.
2011-05-17 Adam Roben <aroben@apple.com>
Make run-api-tests work on Windows when there are spaces in the path
Fixes <http://webkit.org/b/60954> REGRESSION (r86511): run-api-tests fails if there are
spaces in the path to TestWebKitAPI.exe
Reviewed by David Levin.
* Scripts/run-api-tests:
(runTestTool): Use the "direct object" form of system() to avoid having the path to
TestWebKitAPI.exe be split by the shell.
2011-05-16 Ian Henderson <ianh@apple.com>
Reviewed by Joseph Pecoraro.
Page::goToItem doesn't work while loading is deferred
https://bugs.webkit.org/show_bug.cgi?id=60412
Add setDefersLoading and goBack methods to LayoutTestController. We
need to use goBack() instead of history.back() because the latter goes
through NavigationScheduler, hence doesn't exhibit the bug.
* DumpRenderTree/LayoutTestController.cpp:
(goBackCallback):
(setDefersLoadingCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::goBack):
(LayoutTestController::setDefersLoading):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::goBack):
(LayoutTestController::setDefersLoading):
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::goBack):
(LayoutTestController::setDefersLoading):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::goBack):
(LayoutTestController::setDefersLoading):
2011-05-16 Sam Weinig <sam@webkit.org>
Reviewed by Anders Carlsson.
TestWebKitAPI should build with clang if it can
https://bugs.webkit.org/show_bug.cgi?id=60918
* TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
Update CompilerVersion.xcconfig to match others.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
Add CompilerVersion.xcconfig to the project.
* TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
(TestWebKitAPI::didFailProvisionalLoadWithErrorForFrame):
* TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
(TestWebKitAPI::didSameDocumentNavigationForFrame):
Add some casts to quiet warnings from clang.
2011-05-16 Dirk Pranke <dpranke@chromium.org>
Reviewed by David Levin.
add dpranke as a reviewer
https://bugs.webkit.org/show_bug.cgi?id=60919
* Scripts/webkitpy/common/config/committers.py:
2011-05-16 Robert Hogan <robert@webkit.org>
Reviewed by Anders Carlsson.
[Gtk] plugins/get-url-notify-with-url-that-fails-to-load.html crashes on buildbot
https://bugs.webkit.org/show_bug.cgi?id=60838
The unix test plugin needs to call the test's NPP_URLNotify
rather than the browsers.
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_url_notify):
2011-05-16 Brent Fulgham <bfulgham@webkit.org>
Rubber-stamped by Adam Roben.
Adding myself as a reviewer.
* Scripts/webkitpy/common/config/committers.py:
2011-05-16 Sam Weinig <sam@webkit.org>
Fix typo pointed out by Dave Levin.
* TestWebKitAPI/PlatformUtilities.h:
(TestWebKitAPI::Util::assertWKStringEqual):
Strig -> String.
2011-05-16 Sam Weinig <sam@webkit.org>
Reviewed by David Levin.
Convert api tester over to using gtest expectations directly
https://bugs.webkit.org/show_bug.cgi?id=60862
* TestWebKitAPI/PlatformUtilities.cpp:
(TestWebKitAPI::Util::toSTD):
* TestWebKitAPI/PlatformUtilities.h:
(TestWebKitAPI::Util::assertWKStrigEqual):
Add convenience macro to compare WK2 strings. Add some overloads of toSTD to make
the implementation of the macro simpler.
* TestWebKitAPI/Test.h:
Remove TEST_ASSERT forwarder.
[Test changes elided]
2011-05-16 David Kilzer <ddkilzer@apple.com>
<http://webkit.org/b/60913> C++ exceptions should not be enabled when building with llvm-gcc-4.2
<rdar://problem/9446430>
Reviewed by Mark Rowe.
* MiniBrowser/Configurations/Base.xcconfig: Fixed typo.
* TestWebKitAPI/Configurations/Base.xcconfig: Ditto.
* WebKitTestRunner/Configurations/Base.xcconfig: Ditto.
2011-05-16 Jon Lee <jonlee@apple.com>
Reviewed by Simon Fraser.
Can't horizontally scroll iframes and overflow because wheel events are always accepted
https://bugs.webkit.org/show_bug.cgi?id=60779
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage): Set new default method to nil.
2011-05-16 Ademar de Souza Reis Jr. <ademar.reis@openbossa.org>
Reviewed by Csaba Osztrogonác.
[Qt] Tools.pro misses include(common.pri)
https://bugs.webkit.org/show_bug.cgi?id=60883
It includes features.pri, which depends on common.pri (the build is
not broken today with the default options, but this is the right
thing to do).
* Tools.pro:
2011-05-15 Dominic Cooney <dominicc@chromium.org>
Unreviewed: adding myself to committers.py.
* Scripts/webkitpy/common/config/committers.py:
2011-05-15 Daniel Bates <dbates@webkit.org>
Reviewed by Chris Jerdonek.
svn-apply can't handle Git diff that only changes executable bit
https://bugs.webkit.org/show_bug.cgi?id=60848
Fixes an issue where svn-{apply, unapply} fail to handle a Git diff for a file
that only has an executable bit change.
* Scripts/VCSUtils.pm:
- Modified regex $chunkRangeRegEx to match chunk range with ending text, say: @@ -2,6 +2,18 @@ foo().
(parseDiff): Modified to count the number of text chunks in the diff. This should also help towards
fixing <https://bugs.webkit.org/show_bug.cgi?id=29684>.
* Scripts/svn-apply:
(patch): Only use standard patch tool if the patch is non-empty and for a text file (i.e. not a binary patch).
* Scripts/svn-unapply:
(patch): Ditto.
* Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
- Updated expected results for test cases to account for numTextChunks.
- Fixed malformed chunk range in test cases: "Git: simple", and "Git: Git diff followed by SVN diff".
- Added test case "Git: file that only has an executable bit change".
* Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl:
- Updated expected results for test cases to account for numTextChunks.
2011-05-15 Sam Weinig <sam@webkit.org>
Fix windows build.
* TestWebKitAPI/win/main.cpp:
(main):
2011-05-15 Sam Weinig <sam@webkit.org>
Reviewed by David Levin.
Make API tester more gtest friendly
https://bugs.webkit.org/show_bug.cgi?id=60859
* Scripts/run-api-tests:
Numerous changes:
- Remove logic to run tests one at a time.
- Remove logic to build up a list of tests, instead just pass --gtest_list_tests to
the test runner and allow gtest to print it for us.
- Forward stdout from the test tool to stdout, even in non-verbose runs.
- Run test tool from the current working directory instead of the root webkit
directory.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
Move plist to be with other bundle sources.
* TestWebKitAPI/TestsController.cpp:
(TestWebKitAPI::TestsController::TestsController):
(TestWebKitAPI::TestsController::run):
* TestWebKitAPI/TestsController.h:
* TestWebKitAPI/mac/main.mm:
Simplify TestController down to a single run function. We should consider removing the class
entirely at some point.
2011-05-15 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r86504.
http://trac.webkit.org/changeset/86504
https://bugs.webkit.org/show_bug.cgi?id=60853
"Broke Qt EventSender in editing/selection tests" (Requested
by mwenge on #webkit).
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::WebPage):
(WebCore::DumpRenderTree::DumpRenderTree):
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
(WebCore::DumpRenderTree::initJSObjects):
* DumpRenderTree/qt/DumpRenderTreeQt.h:
(WebCore::DumpRenderTree::eventSender):
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_new_instance):
(webkit_test_plugin_set_window):
(webkit_test_plugin_handle_event):
2011-05-12 Robert Hogan <robert@webkit.org>
Reviewed by Benjamin Poulain.
[Qt] fix http/tests/plugins/plugin-document-has-focus.html
The support for this test added to the unix test plugin here
may allow other platforms to pass it (nearly everyone
skips it). On Qt it required a bit of trickiness with
the page's EventSender object to get it working fully though,
so only unskipping Qt here.
https://bugs.webkit.org/show_bug.cgi?id=60722
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::WebPage):
(WebCore::WebPage::eventSender):
(WebCore::DumpRenderTree::DumpRenderTree):
(WebCore::DumpRenderTree::initJSObjects):
(WebCore::DumpRenderTree::createWindow):
* DumpRenderTree/qt/DumpRenderTreeQt.h:
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_new_instance):
(webkit_test_plugin_set_window):
(webkit_test_plugin_handle_event):
2011-05-14 Arno Renevier <arno@renevier.net>
Reviewed by Martin Robinson.
make run-gtk-tests executable
https://bugs.webkit.org/show_bug.cgi?id=59278
* Scripts/run-gtk-tests:
2011-05-14 Anders Carlsson <andersca@apple.com>
Reviewed by Kevin Decker.
NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
https://bugs.webkit.org/show_bug.cgi?id=60823
<rdar://problem/9430386>
Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
* DumpRenderTree/DumpRenderTree.gypi:
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
(GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
(GetURLNotifyWithURLThatFailsToLoad::NPP_New):
(GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* GNUmakefile.am:
2011-05-14 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r86476.
http://trac.webkit.org/changeset/86476
https://bugs.webkit.org/show_bug.cgi?id=60832
It made plugins/get-url-notify-with-url-that-fails-to-
load.html crash on GTK (Requested by Ossy on #webkit).
* DumpRenderTree/DumpRenderTree.gypi:
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Removed.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* GNUmakefile.am:
2011-05-13 Anders Carlsson <andersca@apple.com>
Reviewed by Sam Weinig.
NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
https://bugs.webkit.org/show_bug.cgi?id=60823
<rdar://problem/9430386>
Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
* DumpRenderTree/DumpRenderTree.gypi:
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
(GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
(GetURLNotifyWithURLThatFailsToLoad::NPP_New):
(GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* GNUmakefile.am:
2011-05-13 Anders Carlsson <andersca@apple.com>
Build fix.
* DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp:
2011-05-13 Anders Carlsson <andersca@apple.com>
Reviewed by Sam Weinig.
NPN_GetURL with a javascript: URL doesn't send back the result in the stream
https://bugs.webkit.org/show_bug.cgi?id=60810
Add a test that calls NPN_GetURL on a javascript: URL and checks that it gets back the expected result.
* DumpRenderTree/DumpRenderTree.gypi:
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* GNUmakefile.am:
Add files.
* DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
* DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
Add new NPN and NPP wrappers and sort them according to the order they appear in npfunctions.h
* DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp: Added.
(GetURLWithJavaScriptURL::GetURLWithJavaScriptURL):
(GetURLWithJavaScriptURL::NPP_New):
(GetURLWithJavaScriptURL::NPP_NewStream):
(GetURLWithJavaScriptURL::NPP_DestroyStream):
(GetURLWithJavaScriptURL::NPP_WriteReady):
(GetURLWithJavaScriptURL::NPP_Write):
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(NPP_NewStream):
(NPP_WriteReady):
(NPP_Write):
(NPP_URLNotify):
call the PluginTest NPP wrappers.
2011-05-13 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r86447.
http://trac.webkit.org/changeset/86447
https://bugs.webkit.org/show_bug.cgi?id=60809
"Broke some uses of EventSender object on Qt" (Requested by
mwenge on #webkit).
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::WebPage):
(WebCore::DumpRenderTree::DumpRenderTree):
(WebCore::DumpRenderTree::initJSObjects):
(WebCore::DumpRenderTree::createWindow):
* DumpRenderTree/qt/DumpRenderTreeQt.h:
(WebCore::DumpRenderTree::eventSender):
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_new_instance):
(webkit_test_plugin_set_window):
(webkit_test_plugin_handle_event):
2011-05-13 Robert Hogan <robert@webkit.org>
Reviewed by Benjamin Poulain.
[Qt] fix http/tests/plugins/plugin-document-has-focus.html
The support for this test added to the unix test plugin here
may allow other platforms to pass it (nearly everyone
skips it). On Qt it required a bit of trickiness with
the page's EventSender object to get it working fully though,
so only unskipping Qt here.
https://bugs.webkit.org/show_bug.cgi?id=60722
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::WebPage):
(WebCore::WebPage::eventSender):
(WebCore::DumpRenderTree::DumpRenderTree):
(WebCore::DumpRenderTree::initJSObjects):
(WebCore::DumpRenderTree::createWindow):
* DumpRenderTree/qt/DumpRenderTreeQt.h:
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(webkit_test_plugin_new_instance):
(webkit_test_plugin_set_window):
(webkit_test_plugin_handle_event):
2011-05-13 Adam Roben <aroben@apple.com>
Exclude the Leopard NRWT bot from the core builder set
Fixes <http://webkit.org/b/60392> Leopard Intel Release (NRWT) shouldn't be a core builder
Reviewed by Darin Adler.
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
(BuildBot.__init__): Tighten up the Leopard regexp to only allow the non-NRWT Leopard bots
in the core builder set.
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
(BuildBotTest.test_builder_name_regexps): Updated to match the current set of builders and
regexps.
2011-05-13 Adam Roben <aroben@apple.com>
Group all the NRWT bots together
This also gets the new NRWT bots out of the middle of Apple's bots. Since these new NRWT
bots are experimental, it seems best not to clutter up the non-experimental bots with
experimental ones.
Fixes <http://webkit.org/b/60769> New "(NRWT)" bots are right in the middle of Apple's bots
on build.webkit.org
Reviewed by Dan Bernstein.
* BuildSlaveSupport/build.webkit.org-config/config.json: Moved the "(NRWT)" bots next to the
Chromium testers, which also use new-run-webkit-tests.
2011-05-13 Adam Roben <aroben@apple.com>
Add a plugin test that dumps the plugin window's rect
Part of the test for <http://webkit.org/b/60194> <rdar://problem/9383760> REGRESSION
(r79167): Windowed plugins in Google Reader don't move when the article list is scrolled
Reviewed by Dan Bernstein.
* DumpRenderTree/TestNetscapePlugIn/Tests/win/DumpWindowRect.cpp: Added.
(DumpWindowRect::DumpWindowRect): Just call up to the base class.
(DumpWindowRect::performWindowGeometryTest): Find our window rect relative to the test
harness window and log it.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added DumpWindowRect.
2011-05-13 Adam Roben <aroben@apple.com>
Move code required for most tests of plugin window geometry into a base class
WebKit2's asynchronous plugin window positioning makes testing window geometry tricky. By
having all the trickiness in a shared base class, writing new window geometry tests will be
much easier.
Fixes <http://webkit.org/b/60776> It's hard to write tests that check plugin window geometry
Reviewed by Anders Carlsson.
* DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
(WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Changed to inherit from the new
WindowGeometryTest base class. Most of the code from this class was moved to the new base
class.
(WindowRegionIsSetToClipRect::performWindowGeometryTest): Renamed from checkWindowRegion.
This is called by the base class when it is safe to check the window's geometry. Note that
the m_didCheckWindowRegion logic was removed completely; it has never been needed and was
just vestigial code from an earlier version of this test.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new
WindowGeometryTest class.
* DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.cpp: Added. Code came from
WindowRegionIsSetToClipRect.
(WindowGeometryTest::finishTest): This is the only change from WindowRegionIsSetToClipRect.
We now call the new virtual performWindowGeometryTest function instead of checkWindowRegion.
Subclasses can implement that function to perform their own tests.
* DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.h: Added.
2011-05-13 Zoltan Herczeg <zherczeg@webkit.org>
Reviewed by Csaba Osztrogonác.
Missing 2nd #EOF if WebProcess is killed.
https://bugs.webkit.org/show_bug.cgi?id=60761
If the WebProcess is killed, only one #EOF is yielded
by WebKitTestRunner. However, run-webkit-test wait for
a second #EOF, which signals the finishing of pixel dump.
Since this never happens, run-webkit-test waits forever.
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::dump):
(WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
* WebKitTestRunner/TestInvocation.h:
2011-05-13 Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[NRWT] chromium-linux tests shouldn't run on other platforms
https://bugs.webkit.org/show_bug.cgi?id=60162
Remove _tests_for_other_platforms defs because r82753
implemented it properly in the baseclass in webkit.py.
* Scripts/webkitpy/layout_tests/port/gtk.py:
* Scripts/webkitpy/layout_tests/port/qt.py:
* Scripts/webkitpy/layout_tests/port/win.py:
2011-05-12 Ryuan Choi <ryuan.choi@samsung.com>
Reviewed by Antonio Gomes.
[EFL] Remove unused option of EWeblauncher
https://bugs.webkit.org/show_bug.cgi?id=60747
Remove sudo-workaround option.
This option is not used and make other option wrong.
* EWebLauncher/main.c:
2011-05-12 Lucas Forschler <lforschler@apple.com>
Reviewed by Stephanie Lewis.
Fix config.json to address missing WK2 builder/tester config.
Fix naming of machines (xserve not macpro)
* BuildSlaveSupport/build.webkit.org-config/config.json:
2011-05-12 Lucas Forschler <lforschler@apple.com>
Reviewed by Stephanie Lewis.
Add SnowLeopard Debug builder and Test bot.
apple-xserve-8 = SL Debug Tester.
apple-xserve-9 = SL Debug Builder.
* BuildSlaveSupport/build.webkit.org-config/config.json:
2011-05-12 Adam Barth <abarth@webkit.org>
This patch disables sheriffbot's bug postings. He's just been way too
spammy recently due to test flakiness.
* Scripts/webkitpy/tool/commands/sheriffbot.py:
* Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
2011-05-11 Brian Weinstein <bweinstein@apple.com>
Reviewed by Anders Carlsson.
After relaunching an unresponsive web process, web page stops responding to mouse movement
https://bugs.webkit.org/show_bug.cgi?id=60669
<rdar://problem/9409934>
This test follows the steps needed to reproduce this bug:
1) Hang the web process
2) While the web process is hung, move the mouse (so WebPageProxy thinks it is processing a mouse move event)
3) Kill and restart the WebProcess
4) Move the mouse
We are testing that we receive mouse move events after the WebProcess has restarted.
* TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp: Added.
(TestWebKitAPI::didFinishLoadForFrame): Set a flag that we finished loading.
(TestWebKitAPI::processDidBecomeUnresponsive): Kill the web process once we become unresponsive.
(TestWebKitAPI::processDidCrash): Reload the page when the process crashes, which restarts the web process.
(TestWebKitAPI::setPageLoaderClient):
(TestWebKitAPI::TEST):
* TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash_Bundle.cpp: Added.
(TestWebKitAPI::MouseMoveAfterCrashTest::MouseMoveAfterCrashTest):
(TestWebKitAPI::MouseMoveAfterCrashTest::didReceiveMessage):
* TestWebKitAPI/Tests/WebKit2/mouse-move-listener.html: Added.
* TestWebKitAPI/PlatformWebView.h:
* TestWebKitAPI/mac/PlatformWebViewMac.mm:
(TestWebKitAPI::PlatformWebView::simulateMouseMove): Sends the messages that simulate a mouse move.
* TestWebKitAPI/win/PlatformWebViewWin.cpp:
(TestWebKitAPI::PlatformWebView::simulateMouseMove): Ditto.
Add new files.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
* TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
* TestWebKitAPI/win/copy-resources.cmd:
2011-05-12 Tomasz Morawski <t.morawski@samsung.com>
Reviewed by Antonio Gomes.
[EFL] Added initialization/shutdown for ecore_file library
https://bugs.webkit.org/show_bug.cgi?id=60457
Added initialization and shuthown for ecore_file library and replaced
realpath function to ecore_file_realpath.
* EWebLauncher/main.c:
(quit):
(findThemePath):
(main):
2011-05-12 Young Han Lee <joybro@company100.net>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Implement layoutTestController.layerTreeAsText()
https://bugs.webkit.org/show_bug.cgi?id=60367
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::layerTreeAsText):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-05-11 Eric Seidel <eric@webkit.org>
Reviewed by David Levin.
Teach validate-committer-lists how to validate committers bugzilla_email()
https://bugs.webkit.org/show_bug.cgi?id=60651
It turns out we currently have 9 committers with invalid values for their
bugzilla_email field (meaning we don't know what email is their bugzilla account).
"Alexander Kellett" <lypanov@mac.com>
"Andre Boule" <aboule@apple.com>
"Cameron McCormack" <cam@webkit.org>
"Chris Petersen" <cpetersen@apple.com>
"Eli Fidler" <eli@staikos.net>
"Matt Lilek" <webkit@mattlilek.com>
"Pierre d'Herbemont" <pdherbemont@free.fr>
"Scott Violet" <sky@chromium.org>
"Adam Treat" <treat@kde.org>
Most of those haven't committed in a very long time, so it's not suprising
that we haven't noticed that their bugzilla emails are bogus.
* Scripts/validate-committer-lists:
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
2011-05-11 Dmitry Lomov <dslomov@google.com>
Reviewed by David Levin.
https://bugs.webkit.org/show_bug.cgi?id=59561
Switch TestWebKitAPI to GTest
Fixes Makefiles and WinCairo build as well
* Makefile:
* Scripts/build-api-tests:
* Scripts/run-api-tests:
* TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
* TestWebKitAPI/Test.h:
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
(TestWebKitAPI::TEST):
* TestWebKitAPI/TestsController.cpp:
(TestWebKitAPI::TestsController::TestsController):
(TestWebKitAPI::TestsController::dumpTestNames):
(TestWebKitAPI::TestsController::runTestNamed):
(TestWebKitAPI::TestsController::runAllTests):
* TestWebKitAPI/TestsController.h:
* TestWebKitAPI/mac/main.mm:
(main):
2011-05-11 Eric Seidel <eric@webkit.org>
Reviewed by Alexey Proskuryakov.
Add Luke Macpherson to committers.py as a Contributor
https://bugs.webkit.org/show_bug.cgi?id=60639
* Scripts/webkitpy/common/config/committers.py:
2011-05-11 Noam Rosenthal <noam.rosenthal@nokia.com>
Reviewed by Kenneth Rohde Christiansen.
[Texmap][Qt] Upstream texture-mapper changes from Qt's WebKit2 branch
https://bugs.webkit.org/show_bug.cgi?id=60439
Patch 4/12: Update the build script to enable 3D rendering by default for Qt.
* Scripts/build-webkit:
2011-05-11 Brian Weinstein <bweinstein@apple.com>
Reviewed by Adam Roben.
REGRESSION (r86161): WKPreferences API test failing on Windows
https://bugs.webkit.org/show_bug.cgi?id=60658
* TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
(TestWebKitAPI::TEST):
2011-05-11 Alice Boxhall <aboxhall@chromium.org>
Reviewed by Ojan Vafai.
Modify jsonresults_unittest.py to use a dict format for its test data, and modify jsonresults.py to flatten hierarchical directory structures in input JSON.
https://bugs.webkit.org/show_bug.cgi?id=60521
First steps towards making the results JSON all hierarchical.
* TestResultServer/model/jsonresults.py:
* TestResultServer/model/jsonresults_unittest.py:
2011-05-11 Kevin Ollivier <kevino@theolliviers.com>
[wx] Build fix, adding new LayoutTestController method stubs.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::localStorageDiskUsageForOrigin):
(LayoutTestController::applicationCacheDiskUsageForOrigin):
2011-05-11 Jessie Berlin <jberlin@apple.com>
Reviewed by Darin Adler.
[WebKit2] exceededDatabaseQuota function on WKPageUIClient needs to provide information
about the usage for both the origin overall and the database in question.
https://bugs.webkit.org/show_bug.cgi?id=60573
* WebKitTestRunner/TestController.cpp:
(WTR::exceededDatabaseQuota):
2011-05-10 Stephen White <senorblanco@chromium.org>
Reviewed by David Levin.
Adding myself as a reviewer.
https://bugs.webkit.org/show_bug.cgi?id=60587
* Scripts/webkitpy/common/config/committers.py:
2011-05-11 Eric Seidel <eric@webkit.org>
Reviewed by Csaba Osztrogonác.
queues.webkit.org should show 1000+ instead of 1000 when we hit the query limit
https://bugs.webkit.org/show_bug.cgi?id=60614
AppEngine has a result limit of 1000. We hit that when trying
to find out how many patches the style-queue (and other queues)
have passed in the last 30 days. So lets display 1000+ instead
of 1000 to avoid confusion. (Ossy was asking about this in #webkit tonight.)
I also made the status pages show "never" instead of just "ago" when we
haven't seen a pass or boot message in the last 1000 messages.
* QueueStatusServer/handlers/queuestatus.py:
* QueueStatusServer/templates/queuestatus.html:
2011-05-11 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
sheriffbot shouldn't spam bugs when it can't figure out who to blame
https://bugs.webkit.org/show_bug.cgi?id=60613
This patch limits the number of bugs sheriffbot will spam to three.
Ideally, we'd come up with some smarter huerstics, but hopefully this
will cut down on the bugmail spam for slow bots.
* Scripts/webkitpy/tool/commands/sheriffbot.py:
2011-05-10 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
EWS bots should be robust against test-webkitpy hangs
https://bugs.webkit.org/show_bug.cgi?id=60579
Added a killAfterTimeout subroutine, which takes a child_pid
and timeout, and will setup a watchdog process to kill that child
after the timeout. killAfterTimeout will waitpid on the child.
* EWSTools/start-queue.sh:
2011-05-10 MORITA Hajime <morrita@google.com>
Unreviewed, moved myself to the reviewers list.
* Scripts/webkitpy/common/config/committers.py:
2011-05-10 Anton D'Auria <adauria@apple.com>
Reviewed by David Levin.
StorageTracker should report actual local storage usage on disk
https://bugs.webkit.org/show_bug.cgi?id=60558
* DumpRenderTree/LayoutTestController.cpp:
(localStorageDiskUsageForOriginCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::localStorageDiskUsageForOrigin):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::localStorageDiskUsageForOrigin):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::localStorageDiskUsageForOrigin):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::localStorageDiskUsageForOrigin):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::localStorageDiskUsageForOrigin):
2011-05-10 Anton D'Auria <adauria@apple.com>
Reviewed by Alexey Proskuryakov.
Add ApplicationCache API to get disk usage per origin
https://bugs.webkit.org/show_bug.cgi?id=57127
Added test that loads two frames with ~15Kb image, checks usage,
deletes app cache, verifies usage is 0.
* DumpRenderTree/LayoutTestController.cpp:
(applicationCacheDiskUsageForOriginCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
(LayoutTestController::LayoutTestController):
(LayoutTestController::applicationCacheDiskUsageForOrigin):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
(LayoutTestController::applicationCacheDiskUsageForOrigin):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::applicationCacheDiskUsageForOrigin):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
(LayoutTestController::applicationCacheDiskUsageForOrigin):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
(LayoutTestController::applicationCacheDiskUsageForOrigin):
2011-05-10 Lucas De Marchi <lucas.demarchi@profusion.mobi>
Unreviewed, adding Rafael Antognolli as a committer.
* Scripts/webkitpy/common/config/committers.py:
2011-05-10 Chris Evans <cevans@chromium.org>
Reviewed by Adam Barth.
Add WebCore::Setting to block displaying and/or running insecure content on secure pages
https://bugs.webkit.org/show_bug.cgi?id=58378
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setAllowDisplayOfInsecureContent):
(LayoutTestController::setAllowRunningOfInsecureContent):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::reset):
(WebPreferences::applyTo):
* DumpRenderTree/chromium/WebPreferences.h:
Add plumbing for new settings to Chromium DRT.
2011-05-10 Enrica Casucci <enrica@apple.com>
Reviewed by Beth Dakin.
Added my name to the reviewers list and removed my name from
to committer-only list.
* Scripts/webkitpy/common/config/committers.py:
2011-05-10 David Levin <levin@chromium.org>
Reviewed by Shinichiro Hamaji.
check-webkit-style shouldn't filter the errors using the modified lines for the xml and test expectations checkers.
https://bugs.webkit.org/show_bug.cgi?id=60466
* Scripts/webkitpy/style/checker.py: (unrelated) style fix.
* Scripts/webkitpy/style/checker_unittest.py: Remove the file_path check since
that property was removed.
* Scripts/webkitpy/style/checkers/test_expectations.py: Turn off the line
filtering for errors.
* Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add testing to
verify that the line filtering is turned off.
* Scripts/webkitpy/style/checkers/xml.py: Turn off the line filtering for errors
and remove the unused file_path parameter.
* Scripts/webkitpy/style/checkers/xml_unittest.py: Add testing to verify that the
line filtering is turned off (and fix test_no_error).
* Scripts/webkitpy/style/error_handlers.py: Added a way to turn off the filtering
of errors based on the lines changed in the patch.
* Scripts/webkitpy/style/error_handlers_unittest.py: Added a test for turning off
the filtering.
2011-05-10 Adam Roben <aroben@apple.com>
Update some plugin tests to work with WebKit2's asynchronous windowed plugin positioning on Windows
Fixes <http://webkit.org/b/60442> <rdar://problem/9403200> REGRESSION (r85795):
platform/win/plugins/window-geometry-initialized-before-set-window.html and
platform/win/plugins/window-region-is-set-to-clip-rect.html failing on Windows 7 Release
(WebKit2 Tests)
Reviewed by Anders Carlsson.
* DumpRenderTree/TestNetscapePlugIn/Tests/win/NPNInvalidateRectInvalidatesWindow.cpp:
(NPNInvalidateRectInvalidatesWindow::NPP_SetWindow): Moved some code from here to the new
WindowedPluginTest::testHarnessWindow function.
* DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowGeometryInitializedBeforeSetWindow.cpp:
(WindowGeometryInitializedBeforeSetWindow::NPP_SetWindow): Only check the window's size, not
its position. In WebKit2, the position won't have been set yet, so we can't test it. But it
turns out only the size mattered for <http://webkit.org/b/46716>, which this test was
created for.
* DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
(WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Converted to be a
WindowedPluginTest, and initialized new members.
(WindowRegionIsSetToClipRect::startTest): Make our plugin window visible, and set things up
to ensure we get a WM_WINDOWPOSCHANGED message about it becoming visible.
(WindowRegionIsSetToClipRect::finishTest): Check that the window region is what we expect,
then clean up and tell the test harness we're done.
(WindowRegionIsSetToClipRect::checkWindowRegion): Moved code to check the window region here
from NPP_SetWindow. Changed to use the fancy new log() function instead of pluginLog().
(WindowRegionIsSetToClipRect::showTestHarnessWindowIfNeeded): Show the test harness's window
if it is hidden. This is needed to ensure we get a WM_WINDOWPOSCHANGED message when our own
window becomes visible.
(WindowRegionIsSetToClipRect::hideTestHarnessWindowIfNeeded): Hide the test harness's window
if it was hidden at the start of the test.
(WindowRegionIsSetToClipRect::wndProc): When our timer fires, tell the UI process to paint.
This should cause our window to become visible and our window region to get set. When we get
a WM_WINDOWPOSCHANGED message, we can finally finish the test.
(WindowRegionIsSetToClipRect::NPP_GetValue): Return our script object.
(WindowRegionIsSetToClipRect::ScriptObject::hasMethod): We only have one method:
"startTest".
(WindowRegionIsSetToClipRect::ScriptObject::invoke): Call through to the PluginTest object.
* DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.cpp:
(WindowedPluginTest::testHarnessWindow):
* DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.h:
Added this helper function. Code came from NPNInvalidateRectInvalidatesWindow.
2011-05-10 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Add new qt-4.8 platform for layout testing
https://bugs.webkit.org/show_bug.cgi?id=60544
* Scripts/old-run-webkit-tests:
* Scripts/webkitdirs.pm:
2011-05-09 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r86108.
http://trac.webkit.org/changeset/86108
https://bugs.webkit.org/show_bug.cgi?id=60533
Broke makefile based builds and Cairo (Requested by weinig on
#webkit).
* Scripts/build-api-tests:
* Scripts/run-api-tests:
* TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
* TestWebKitAPI/Test.h:
(TestWebKitAPI::Test::~Test):
(TestWebKitAPI::Test::name):
(TestWebKitAPI::Test::Register::Register):
(TestWebKitAPI::Test::Register::create):
(TestWebKitAPI::Test::Test):
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
(TestWebKitAPI::TEST):
* TestWebKitAPI/TestsController.cpp:
(TestWebKitAPI::TestsController::TestsController):
(TestWebKitAPI::TestsController::dumpTestNames):
(TestWebKitAPI::TestsController::runTestNamed):
(TestWebKitAPI::TestsController::testFailed):
(TestWebKitAPI::TestsController::registerCreateTestFunction):
* TestWebKitAPI/TestsController.h:
* TestWebKitAPI/mac/main.mm:
(main):
2011-05-09 Hayato Ito <hayato@chromium.org>
Reviewed by Ojan Vafai.
Skip reftests if pixel tests are disabled.
https://bugs.webkit.org/show_bug.cgi?id=59078
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2011-05-09 Chang Shu <cshu@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt] DumpRenderTreeQt needs an implementation of unmarkText
https://bugs.webkit.org/show_bug.cgi?id=60499
The implementation is similar to gtk.
* DumpRenderTree/qt/TextInputControllerQt.cpp:
(TextInputController::unmarkText):
* DumpRenderTree/qt/TextInputControllerQt.h:
2011-05-09 James Robinson <jamesr@chromium.org>
Reviewed by Eric Seidel.
[chromium] Teach new-run-webkit-tests to follow symlinks when trying to determine the arch of DumpRenderTree on linux
https://bugs.webkit.org/show_bug.cgi?id=60520
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2011-05-09 Dmitry Lomov <dslomov@google.com>
Reviewed by Sam Weinig.
Switch TestWebKitAPI to GTest
https://bugs.webkit.org/show_bug.cgi?id=59561
* Scripts/build-api-tests:
* Scripts/run-api-tests:
* TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
* TestWebKitAPI/Test.h:
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
(TestWebKitAPI::TEST):
* TestWebKitAPI/TestsController.cpp:
(TestWebKitAPI::TestsController::TestsController):
(TestWebKitAPI::TestsController::dumpTestNames):
(TestWebKitAPI::TestsController::runTestNamed):
(TestWebKitAPI::TestsController::runAllTests):
* TestWebKitAPI/TestsController.h:
* TestWebKitAPI/mac/main.mm:
(main):
2011-05-09 Igor Oliveira <igor.oliveira@openbossa.org>
Reviewed by Andreas Kling.
[Qt] DRT eventSender.addTouchPoint method is not setting the start position
https://bugs.webkit.org/show_bug.cgi?id=60487
When adding a new touch point, the start position of this point needs to be set.
* DumpRenderTree/qt/EventSenderQt.cpp:
(EventSender::addTouchPoint):
2011-05-09 Robert Hogan <robert@webkit.org>
Reviewed by Adam Roben.
https://bugs.webkit.org/show_bug.cgi?id=57137
Unskip plugins/embed-prefers-plugins-for-images.html
Add image/png mime-type to test netscape plugin on Unix.
* DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
(NP_GetMIMEDescription):
2011-05-09 Alice Boxhall <aboxhall@chromium.org>
Reviewed by Ojan Vafai.
Pull jsonresults_unittest test cases out into separate test methods.
https://bugs.webkit.org/show_bug.cgi?id=60418
* TestResultServer/model/jsonresults_unittest.py:
2011-05-09 Mahesh Kulkarni <mahesh.kulkarni@nokia.com>
Reviewed by Andreas Kling.
[Qt] fix fast/dom/Geolocation/window-close-crash.html
https://bugs.webkit.org/show_bug.cgi?id=60458
Update all pages with Mock position/error/permission.
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setGeolocationPermission):
(LayoutTestController::setMockGeolocationError):
(LayoutTestController::setMockGeolocationPosition):
2011-05-09 Shishir Agrawal <shishir@chromium.org>
Reviewed by Tony Gentilcore.
Implement Page Visibility API.
https://bugs.webkit.org/show_bug.cgi?id=54181
* DumpRenderTree/LayoutTestController.cpp:
(setPageVisibilityCallback):
(resetPageVisibilityCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
(LayoutTestController::setPageVisibility):
(LayoutTestController::resetPageVisibility):
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::resetPageVisibility):
(LayoutTestController::setPageVisibility):
* DumpRenderTree/chromium/LayoutTestController.h:
2011-05-07 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Testing EWS spins on patches with a large number of failures
https://bugs.webkit.org/show_bug.cgi?id=60441
In cases where the tree has a small number of persistent failures but a
patch has a large number of failures, we weren't believing any results.
That lead to us retrying the patch forever while we waited for
trustworthy results that would never come.
This patch loosens the semantics of unexpected_failures to return a
subset (rather than the exact set) of unexpected failures. In the case
where the tree has a bounded number of failures and the patch has an
unbounded number of failures, we can accurately compute such a subset
and reject the patch.
* Scripts/webkitpy/tool/bot/expectedfailures.py:
* Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
* Scripts/webkitpy/tool/bot/patchanalysistask.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2011-05-07 Joe Mason <jmason@rim.com>
Reviewed by Daniel Bates.
FrameLoader::isProcessingUserGesture is wrong in dispatchWillPerformClientRedirect
https://bugs.webkit.org/show_bug.cgi?id=52211
Dump isUserProcessingGesture in willPerformClientRedirect so that the layout tests can be extended.
(For the Mac and Chromium ports - other ports don't support dumping user gestures in DRT.)
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::willPerformClientRedirect):
(WebViewHost::didStartProvisionalLoad):
* DumpRenderTree/mac/FrameLoadDelegate.mm:
(-[FrameLoadDelegate webView:willPerformClientRedirectToURL:delay:fireDate:forFrame:]):
2011-05-07 Daniel Bates <dbates@webkit.org>
Attempt to fix the build; remove some more references to isTiger(), which was removed in changeset 86007
<http://trac.webkit.org/changeset/86007> (bug #60438).
* Scripts/old-run-webkit-tests:
- Removed reference to isTiger() from openDumpTool(), captureSavedCrashLog(), and
file local variable.
2011-05-07 Pratik Solanki <psolanki@apple.com>
Reviewed by Darin Adler.
Remove references to Tiger in scripts
https://bugs.webkit.org/show_bug.cgi?id=60438
Remove isTiger() and all of the code that depends on it.
* Scripts/build-webkit:
* Scripts/ensure-valid-python:
* Scripts/gdb-safari:
* Scripts/webkitdirs.pm:
2011-05-06 David Grogan <dgrogan@chromium.org>
Reviewed by Eric Seidel.
update link to message about mac 64bit git binaries
https://bugs.webkit.org/show_bug.cgi?id=60411
The original message implied (correctly, at-the-time) that
http://code.google.com/p/git-osx-installer/ doesn't provide 64-bit
binaries. The new message notes that they do. This might prevent someone
else from spending a few minutes confused.
* Scripts/webkitpy/common/checkout/scm/git.py:
2011-05-06 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: crashes in race with crashing DRT
https://bugs.webkit.org/show_bug.cgi?id=60238
This fixes a race where if we wrote the URL to DRT and
DRT crashed immediately, we could get a broken pipe
error back during the write() and not handle it correctly,
causing NRWT to also crash.
* Scripts/webkitpy/layout_tests/port/server_process.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-05-06 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Add a --skip-integrationtests mode to test-webkitpy
https://bugs.webkit.org/show_bug.cgi?id=60354
These two test suites account for 2/3 of test-webkitpy's running time
because they're not actually unit tests. They're integration tests.
This patch adds the ability to skip integration tests by passing a
command line argument to test-webkitpy. The buildbots will continue to
run these integration tests to make sure they don't regress. This
option is for use by developers when working on parts of the code that
are well-covered by unit tests.
* Scripts/webkitpy/common/system/file_lock_integrationtest.py: Copied from Tools/Scripts/webkitpy/common/system/file_lock_unittest.py.
* Scripts/webkitpy/common/system/file_lock_unittest.py: Removed.
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py.
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py: Removed.
* Scripts/webkitpy/test/main.py:
2011-05-06 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
The results.zip generated by the Chromium EWS are too big
https://bugs.webkit.org/show_bug.cgi?id=60352
Currently results.zip created by the Chromium EWS bot contain roughly
1000 tests that we expect to fail. These test results are not helpful
to contributors.
There are several approaches to solving this problem:
1) We could be more selective about which files to include in the zip.
2) We could ask new-run-webkit-tests not to generate results for
expected failures.
3) We could ask new-run-webkit-tests not to run tests that we expect to
fail.
This patch takes approach (3). Currently, we're ignoring the outcome
of failing tests anyway, so there's not much point in actually running
them. As a side benefit, this will speed up the test run by about 6%.
In the future, we might revisit this decision and choose another
approach.
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/common/config/ports_unittest.py:
2011-05-06 Gabor Rapcsanyi <rgabor@webkit.org>
Reviewed by Csaba Osztrogonác.
[Qt] Use qt-arm platform instead of qt-linux on ARM platform.
* Scripts/old-run-webkit-tests:
2011-05-06 Adam Barth <abarth@webkit.org>
Reviewed by Ojan Vafai.
Add an option to NRWT to skip failing tests
https://bugs.webkit.org/show_bug.cgi?id=60295
This option is similar to the way old-run-webkit-tests works. We're
planning to use this option on the testing EWS bots because they don't
use information from failing tests yet.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-05-04 Philippe Normand <pnormand@igalia.com>
Reviewed by Dimitri Glazkov.
Implement LayoutTestController::pseudoShadowId()
https://bugs.webkit.org/show_bug.cgi?id=60034
shadowPseudoId support in the various LayoutTestController
implementations.
* DumpRenderTree/LayoutTestController.cpp:
(shadowPseudoIdCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::shadowPseudoId):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::shadowPseudoId):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::shadowPseudoId):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::shadowPseudoId):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::shadowPseudoId):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::shadowPseudoId):
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::shadowPseudoId):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-05-06 Adam Barth <abarth@webkit.org>
EWS can't reject patches
https://bugs.webkit.org/show_bug.cgi?id=60345
We're supposed to use patch.id() for the attachment ID, silly.
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2011-05-05 Steve Lacey <sjl@chromium.org>
Reviewed by Ojan Vafai.
Fix generation of results html file by passing relative path to scm
methods.
Generate diff images for results html file.
https://bugs.webkit.org/show_bug.cgi?id=60086
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
2011-05-05 Andy Estes <aestes@apple.com>
Reviewed by Daniel Bates.
update-webkit should handle Internal using a different VCS than OpenSource.
https://bugs.webkit.org/show_bug.cgi?id=60310
isGit() and isSVN() cache their results, so Internal will always be
updated using the VCS commands used for OpenSource. If Internal uses a
different VCS, this will fail. Fix this by calling isGitDirectory() and
isSVNDirectory() instead, which do not cache their results.
* Scripts/update-webkit:
2011-05-05 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
nrwt: revamp integration tests incl. tests for servers
https://bugs.webkit.org/show_bug.cgi?id=59985
This change adds the ability to run "integration tests" that
depend on (and affect) the local environment on your machine.
These tests can be run manually by invoking python on the
test file in question and passing the '-i' flag, but are
not run automatically as part of test-webkitpy.
Integration tests are identified by being methods named as
"integration_test_*" instead of "test_*", and we use a
custom TestLoader instead of the default one provided by
unittest.
This change also adds a protected "_default_to_apache" method
to the port object so we can do the right thing when starting
and stopping web servers with less mechanism.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
2011-05-05 Adam Roben <aroben@apple.com>
Robustify DRT's pixel-dumping code against semi-random alpha channel values on Windows
DRT generates a pixel dump on Windows using the WM_PRINT message. This message (just like
most of GDI) doesn't write anything meaningful into the alpha channel. In some cases, the
alpha channel will even get filled with zeros (like when ::GradientFill is used). To
safeguard against this, we fill the alpha channel with constant full opacity after the
WM_PRINT message is handled to ensure we end up with an opaque image. (Note that we can't
just ignore the alpha channel when creating the PNG; Mac's pixel results have an alpha
channel, and we want to be able to compare Mac and Windows pixel results, so Windows pixel
results need an alpha channel, too.)
Fixes <http://webkit.org/b/60300> draws-gradient.html's pixel results can't be correctly
generated due to alpha channel weirdness
Reviewed by Sam Weinig.
* DumpRenderTree/win/PixelDumpSupportWin.cpp:
(makeAlphaChannelOpaque): Added this helper function to fill a bitmap's alpha channel with
constant full opacity.
(createBitmapContextFromWebView): Make the bitmap's alpha channel opaque before wrapping it
in a CGContext or cairo_surface.
2011-05-05 Adam Roben <aroben@apple.com>
Add support for layoutTestController.dumpAsText(true) to WebKitTestRunner
Passing true tells the test harness to dump pixels even though we're not dumping a render
tree.
Fixes <http://webkit.org/b/60298> Calling layoutTestController.dumpAsText(true) in
WebKitTestRunner doesn't result in pixel results being generated
Reviewed by Sam Weinig.
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpPixels
argument to dumpAsText.
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::dump): Updated for dumpAsText change.
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::dumpAsText): Store the dumpPixels argument.
* WebKitTestRunner/InjectedBundle/LayoutTestController.h: Updated for IDL change.
* WebKitTestRunner/win/InjectedBundle.vcproj: Added .idl files as a convenience for Windows
developers.
2011-05-05 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: merge os, version fields in test config
https://bugs.webkit.org/show_bug.cgi?id=60004
This change removes the 'o/s' field from the TestConfiguration
object, and changes the 'MAC', 'WIN', and 'LINUX' modifiers in
the expectations file to macros that expand to the version names
for each platform.
This means that you can once again put any subset of the
operating system versions on a single line, e.g.:
BUGX LEOPARD XP LINUX : fast/html/foo.html = FAIL
This change will break any existing test_expectations.txt
file that uses the "OS-VERSION" macros, which are no longer
needed. It turns out they were never needed, as just VERSION
by itself was sufficient. The test_expectations.txt file should
be updated accordingly prior to landing this change.
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
* Scripts/webkitpy/layout_tests/port/base.py:
2011-05-05 Tony Chang <tony@chromium.org>
Reviewed by Adam Barth.
[chromium] switch from Lucida to free fonts for Hindi and Thai
https://bugs.webkit.org/show_bug.cgi?id=60287
The fonts are provided by the ttf-thai-tlwg and ttf-indic-fonts
packages on Lucid. There are actually multiple Thai fonts in the
package, so I just picked one.
* DumpRenderTree/chromium/TestShellGtk.cpp:
(setupFontconfig):
2011-05-05 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[chromium] google-chrome-64 builds should now fallback to chromium-linux-x86_64 results
https://bugs.webkit.org/show_bug.cgi?id=60290
It used to be that google-chrome-32 was hardy32, google-chrome-64 was
hardy64, chromium-linux was hardy32 and chromium-linux-x86_64 was
lucid64. Because of that, google-chrome-64 used to skip the
chromium-linux-x86_64 results since they were more incorrect.
Now that we're switching to lucid for google-chrome, we want
google-chrome-64 to fall back to chromium-linux-x86_64.
* Scripts/webkitpy/layout_tests/port/google_chrome.py:
2011-05-05 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: merge TestExpectations, TestExpectationsFile classes
https://bugs.webkit.org/show_bug.cgi?id=60002
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
* Scripts/webkitpy/style/checkers/test_expectations.py:
2011-05-05 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
commit-queue should reject patches which fail ewses
https://bugs.webkit.org/show_bug.cgi?id=47534
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2011-05-05 Chang Shu <cshu@webkit.org>
Reviewed by Andreas Kling.
[Qt][WK2] Support PageGroup in QGraphicsWKView/QWKPage constructors
https://bugs.webkit.org/show_bug.cgi?id=59638
1. Call QGraphicsWKView constructor with pageGroupRef.
2. Disable setFontFamily calls as it breaks Qt test fonts.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::resetStateToConsistentValues):
* WebKitTestRunner/qt/PlatformWebViewQt.cpp:
(WTR::WebView::WebView):
(WTR::PlatformWebView::PlatformWebView):
2011-05-05 Joseph Pecoraro <joepeck@webkit.org>
Reviewed by David Kilzer.
Add --only-webkit option to build-webkit
https://bugs.webkit.org/show_bug.cgi?id=60245
* Scripts/build-webkit: If --only-webkit was specified
just build the WebKit project, and not its dependencies.
2011-05-05 Jeff Miller <jeffm@apple.com>
Reviewed by Darin Adler.
Add didNotHandleWheelEvent callback to WKPageUIClient
https://bugs.webkit.org/show_bug.cgi?id=60203
* MiniBrowser/gtk/BrowserWindow.c:
(browserWindowUIClientInit): Stub out didNotHandleWheelEvent callback.
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]): Stub out didNotHandleWheelEvent callback.
* MiniBrowser/win/BrowserView.cpp:
(BrowserView::create): Stub out didNotHandleWheelEvent callback.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage): Stub out didNotHandleWheelEvent callback.
(WTR::TestController::initialize): Stub out didNotHandleWheelEvent callback.
2011-05-04 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Add more missing imports from scm refactor
https://bugs.webkit.org/show_bug.cgi?id=60239
* Scripts/webkitpy/common/checkout/scm/git.py:
* Scripts/webkitpy/common/checkout/scm/svn.py:
2011-05-04 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
sheriff-bot is having more exception trouble
https://bugs.webkit.org/show_bug.cgi?id=60233
A webpage-forced build was causing SB to throw an exception.
We've long had this bug, but we just didn't notice it until now.
In debugging this I found we fetch an absurd amount of data now
that we use buildbot's json (instead of xmlrpc), so I added
the filter=1 parameter to reduce it a little. That required
me to handle the case where filter=1 would strip
result=0 in some cases (it's unclear why it does that).
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2011-05-04 Cris Neckar <cdn@chromium.org>
Unreviewed, Adding Cris Neckar as a committer
* Scripts/webkitpy/common/config/committers.py:
2011-05-04 Siddharth Mathur <siddharth.mathur@nokia.com>
Reviewed by Laszlo Gombos.
[Qt][Wk2][Symbian] Capabilities for Qt MiniBrowser and QtWebProcess on Symbian
https://bugs.webkit.org/show_bug.cgi?id=59462
* MiniBrowser/qt/MiniBrowser.pro: Add Symbian OS capabilities to executable. Also enable data paging.
2011-05-04 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
The testing EWS bots should upload layout-test-results.zip to bugs after failures
https://bugs.webkit.org/show_bug.cgi?id=60223
This required sharing a bit of code between the commit-queue
(which already knew how to do this) and the new EWS testing bots.
In the process I also cleaned up EWS testing a little and
removed some dead code from the commit-queue.
* Scripts/webkitpy/common/config/committervalidator.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
* Scripts/webkitpy/tool/commands/queues.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
2011-05-04 James Kozianski <koz@chromium.org>
Reviewed by Eric Seidel.
Add method to make a Port able to retrieve TestOutputSets from its builders.
https://bugs.webkit.org/show_bug.cgi?id=58211
This will be used in the new rebaseline script - TestOutputs are
potential new baselines.
* Scripts/webkitpy/common/config/urls.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/factory.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-05-04 Eric Seidel <eric@webkit.org>
Reviewed by Ojan Vafai.
sheriff-bot can't parse results.html
https://bugs.webkit.org/show_bug.cgi?id=60196
The problem was that the NRWT bots are currently core
and so sheriff-bot was trying to diagnose failures on
them. It was trying to parse their results.html file
and BeautifulSoup was throwing an exception.
It seems our version of BeautifulSoup was quite old
and unable to handle <script> tags correctly.
This moves us to 3.2 which doesn't seem to have this bug.
* Scripts/webkitpy/thirdparty/BeautifulSoup.py:
2011-05-04 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Make the EWS test logging less verbose
https://bugs.webkit.org/show_bug.cgi?id=60201
This should help us generate less than a hundred MB of logs per hour.
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/common/config/ports_unittest.py:
2011-05-04 Tao Bai <michaelbai@chromium.org>
Reviewed by David Kilzer.
Populate touch-icon url to FrameLoaderClient
https://bugs.webkit.org/show_bug.cgi?id=59143
Disable touch icon by default.
* Scripts/build-webkit:
2011-05-04 Justin Novosad <junov@chromium.org>
Reviewed by Kenneth Russell.
[Chromium] Add gpu acceleration flags to DumpRenderTree
https://bugs.webkit.org/show_bug.cgi?id=60168
Adding support for command line flags
enable-accelerated-2d-canvas an
enable-accelerated-drawing in DumpRenderTree.
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
(TestShell::resetWebSettings):
* DumpRenderTree/chromium/TestShell.h:
(TestShell::setLegacyAccelerated2dCanvasEnabled):
(TestShell::setAcceleratedDrawingEnabled):
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::reset):
(WebPreferences::applyTo):
* DumpRenderTree/chromium/WebPreferences.h:
2011-05-04 Ademar de Souza Reis Jr. <ademar.reis@openbossa.org>
Reviewed by Csaba Osztrogonác.
[Qt] run-qtwebkit-tests misses exec bit
https://bugs.webkit.org/show_bug.cgi?id=60188
Also removing an unecessary import (mostly to let this bugfix goes
through the system, as this is just a chmod).
* Scripts/run-qtwebkit-tests:
2011-05-04 Adam Barth <abarth@webkit.org>
Fix import of ircproxy module to actually import the module. I don't
really understand how this code could have worked before.
* Scripts/webkitpy/common/host.py:
2011-05-04 Siddharth Mathur <siddharth.mathur@nokia.com>
Reviewed by Andreas Kling.
[Qt][Wk2][Symbian] Minor UI facelift for MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=60172
* MiniBrowser/qt/BrowserWindow.cpp: Bring in UI layout changes from QtTestBrowser
(BrowserWindow::BrowserWindow):
2011-05-04 Mihai Parparita <mihaip@chromium.org>
Reviewed by Ojan Vafai.
[NRWT] --new-baseline should put baselines in the directory specified by --additional-platform-directory
https://bugs.webkit.org/show_bug.cgi?id=60088
Makes --new-baseline put new baselines into the first entry in the
baseline search path, including any additional directories passed in via
--additional-platform-directory.
Simplifies baseline_path() by making it return the first entry in the
baseline search path instead of requiring each port to have an
implementation (should not result in any behavior change).
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-05-04 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Fix circular dependency in webkitpy
https://bugs.webkit.org/show_bug.cgi?id=60075
These functions don't belong in bugzilla.py. They only exist there
because they are old. Really, these functions shouldn't be free
functions at all, but that's a patch for another day.
* Scripts/webkitpy/common/checkout/changelog.py:
* Scripts/webkitpy/common/checkout/changelog_unittest.py:
* Scripts/webkitpy/common/checkout/checkout.py:
* Scripts/webkitpy/common/config/urls.py:
* Scripts/webkitpy/common/net/bugzilla/__init__.py:
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
* Scripts/webkitpy/style/checkers/changelog.py:
* Scripts/webkitpy/tool/bot/irc_command.py:
* Scripts/webkitpy/tool/bot/sheriff.py:
* Scripts/webkitpy/tool/commands/upload.py:
2011-05-03 Pratik Solanki <psolanki@apple.com>
Reviewed by Dan Bernstein.
https://bugs.webkit.org/show_bug.cgi?id=60141
Remove reference to libWebKitSystemInterfaceTiger.a from build-webkit
* Scripts/build-webkit:
2011-05-03 Mihai Parparita <mihaip@chromium.org>
Reviewed by Ojan Vafai.
Update rebaseline queue server to handle hierarchical test results
https://bugs.webkit.org/show_bug.cgi?id=60063
Update test result parsing done by the rebaseline queue server to handle
the hierarchical test result output that is generated as of r85254.
* RebaselineQueueServer/templates/builder-queue-edit.html:
2011-05-03 Eric Seidel <eric@webkit.org>
Reviewed by Adam Roben.
webkit-patch post-commits HEAD^.. throws an exception
https://bugs.webkit.org/show_bug.cgi?id=60027
* Scripts/webkitpy/common/checkout/scm/git.py:
2011-05-03 Luiz Agostini <luiz.agostini@openbossa.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Layout tests may timeout if loadFinished receives false
https://bugs.webkit.org/show_bug.cgi?id=60064
When using layoutTestController.queueLoad(), if QWebPage::loadFinished(bool) receives false
then the layout test may timout. It happens that QWebPage::loadFinished() receives false when
a response is ignored, like it should be for 204 responses. It is a problem for creating
tests for 204 responses for example.
This change has been tested and it causes no regressions in layout tests.
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::maybeDump):
2011-05-03 Dirk Pranke <dpranke@chromium.org>
Unreviewed, rolling out r85659.
http://trac.webkit.org/changeset/85659
https://bugs.webkit.org/show_bug.cgi?id=59977
restarting isn't working correctly
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-05-03 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: fix http server startup/shutdown
Previous versions of the code had three problems that made startup
and shutdown flaky. The first is that it would throw exceptions
if it couldn't delete stale log files, which was overly
paranoid. The second is that some of the exceptions weren't
defined properly. The third, and most important, is that it was
using urllib to check if ports were available, which was leaving
sockets in a half-closed state, and keeping ports from being
reused. By switching to raw sockets, we are able to now reliably
restart.
This change also switches the code to using Executives to stop
processes, which will let us delete a bunch of code in the
port/* implementations and fix a weird layering problem in a
subsequent patch.
https://bugs.webkit.org/show_bug.cgi?id=59977
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server_base.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
2011-05-03 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
fix failing printing_unittest on win32
https://bugs.webkit.org/show_bug.cgi?id=59963
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2011-05-03 Adam Roben <aroben@apple.com>
Fix most strict PassOwnPtr violations on Windows
Fixes <http://webkit.org/b/60025> Windows should (almost) build with strict PassOwnPtr
enabled
Reviewed by Anders Carlsson.
* WebKitAPITest/Test.h: Use adoptPtr.
2011-05-03 Caio Marcelo de Oliveira Filho <caio.oliveira@openbossa.org>
Reviewed by Andreas Kling.
Add myself to the committers list.
* Scripts/webkitpy/common/config/committers.py:
2011-05-03 James Robinson <jamesr@chromium.org>
Reviewed by Kenneth Russell.
[chromium] Layout test animations/3d/replace-filling-transform.html failing
https://bugs.webkit.org/show_bug.cgi?id=59839
Add animations/3d to the list of tests run in the chromium-gpu configuration since these tests exercise the compositor.
* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2011-05-03 Lucas De Marchi <lucas.demarchi@profusion.mobi>
Unreviewed, rolling out r85471.
http://trac.webkit.org/changeset/85471
https://bugs.webkit.org/show_bug.cgi?id=59922
The rule of thumb in EFL is that you init in all the places
you use it. So, if in your application you call ecore_evas_*
functions, you have to call ecore_evas_init() first.
* EWebLauncher/main.c: always call *_init() functions when you use
that namespace.
(quit):
(main):
2011-05-02 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
remove no longer used method
https://bugs.webkit.org/show_bug.cgi?id=59981
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2011-05-02 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
add image toggling behavior from old-run-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=59978
Output -diffs.html files. This matches old-run-webkit-tests.
In theory, we could use data urls in the results.html file, but
data urls don't currently have access to local files.
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2011-05-02 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[nrwt] remove some obsolete references to .checksum files
https://bugs.webkit.org/show_bug.cgi?id=59971
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-05-02 Adam Roben <aroben@apple.com>
Take advantage of implicit nullptr_t -> PassOwnPtr conversion
Fixes <http://webkit.org/b/59964> Implicit conversion from std::nullptr_t to PassOwnPtr
doesn't work, but should
Reviewed by Adam Barth.
* DumpRenderTree/chromium/CppBoundClass.h:
* DumpRenderTree/chromium/WebViewHost.cpp:
2011-05-02 Martin Robinson <mrobinson@igalia.com>
Reviewed by Xan Lopez.
[GTK] Compiler warning in Tools/MiniBrowser/main.c
https://bugs.webkit.org/show_bug.cgi?id=59986
Remove a compiler warning.
* MiniBrowser/gtk/main.c: Use full array initializers to prevent some compilers
from complaining.
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Add runtests_unittest.py
https://bugs.webkit.org/show_bug.cgi?id=59955
I forgot to add this file in a previous patch.
* Scripts/webkitpy/tool/steps/runtests_unittest.py: Added.
2011-05-02 Adam Barth <abarth@webkit.org>
Change the testing EWS to log failures explicitly to avoid an infinite loop.
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2011-05-02 Adrienne Walker <enne@google.com>
Unreviewed, fix misspelled contributor name.
* Scripts/webkitpy/common/config/committers.py:
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
cr-linux-ews should run tests
https://bugs.webkit.org/show_bug.cgi?id=59272
This patch turns on testing for the cr-linux-ews. Yay!
* Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
* Scripts/webkitpy/tool/steps/runtests.py:
* Scripts/webkitpy/tool/steps/steps_unittest.py:
2011-05-02 Martin Robinson <mrobinson@igalia.com>
Incorporate a review comment that I mistakenly omitted from my last commit.
* DumpRenderTree/gtk/EventSender.cpp: Use unsigned here instead of unsigned int.
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
EWS should report the list of unexpected failures
https://bugs.webkit.org/show_bug.cgi?id=59923
We cache the results of the first layout test results run on the task
so the EWS can grab them later. Another option is to include them in
the exception that's thrown, but this approach seemed slightly more
aesthetic.
* Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
* Scripts/webkitpy/tool/bot/patchanalysistask.py:
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Add base case for a test-running EWS
https://bugs.webkit.org/show_bug.cgi?id=59920
Later I will subclass this to run the Chromium Linux EWS.
* Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2011-05-02 Anders Carlsson <andersca@apple.com>
Build fix.
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
* WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
2011-05-02 Nico Weber <thakis@chromium.org>
Reviewed by Mihai Parparita.
Add --no-webkit2 to the output of "build-webkit -h"
https://bugs.webkit.org/show_bug.cgi?id=59942
* Scripts/build-webkit:
2011-05-02 Martin Robinson <mrobinson@igalia.com>
Reviewed by Xan Lopez.
[GTK] old-run-webkit-tests should disable the Ubuntu overlay scrollbar
https://bugs.webkit.org/show_bug.cgi?id=59943
Disable the Ubuntu overlay scrollbar while testing, by injecting the
LIBOVERLAY_SCROLLBAR=0 environment variable.
* Scripts/old-run-webkit-tests:
2011-05-02 Martin Robinson <mrobinson@igalia.com>
Reviewed by Eric Seidel.
[GTK] editing/selection/select-out-of-editable.html failing after r83967
https://bugs.webkit.org/show_bug.cgi?id=58667
Do not trigger mouse down events when the EventSender knows that a mouse button is
already down. These type of situations can confuse Xvfb and cause dragging failures.
* DumpRenderTree/gtk/EventSender.cpp:
(mouseDownCallback): Do not trigger another mouse button down event if the mouse button
is already down. This can interfere with drags in Xvfb.
2011-05-02 Zsolt Fehér <h490267@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
DRT sideeffect revealed by r71268.
https://bugs.webkit.org/show_bug.cgi?id=48958
Reset QWebSettings::ZoomTextOnly attribute before all tests.
Call setTextSizeMultiplier(1.0) to reset TextZoomFactor and
PageZoomFactor too. It should be done before resetsettings()
to guarantee resetting QWebSettings::ZoomTextOnly correctly.
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::resetSettings):
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
2011-05-02 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r85469.
http://trac.webkit.org/changeset/85469
https://bugs.webkit.org/show_bug.cgi?id=59928
Broke webkitpy-tests on Windows 7, GTK and Qt. (Requested by
bbandix on #webkit).
* Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Removed.
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2011-05-02 Tomasz Morawski <t.morawski@samsung.com>
Reviewed by Kenneth Rohde Christiansen.
Cleanup EFL initialization/shutdown in the EWebLauncher
https://bugs.webkit.org/show_bug.cgi?id=59922
Removed ecore_evas_init and edje_init form main function due to they
are also called in the ewk_init. Moved ewk_shutdown to quit function.
* EWebLauncher/main.c:
(quit):
(main):
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Add base case for a test-running EWS
https://bugs.webkit.org/show_bug.cgi?id=59920
Later I will subclass this to run the Chromium Linux EWS.
* Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Add the ability to skip python and perl unit tests on EC2
https://bugs.webkit.org/show_bug.cgi?id=59918
* Scripts/webkitpy/common/config/ports.py:
- This is slightly ugly, but it's useful...
* Scripts/webkitpy/common/config/ports_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
* Scripts/webkitpy/tool/steps/runtests.py:
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
webkit-patch --non-interactive is incompatible with new-run-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=59919
new-run-webkit-tests doesn't support --wait-for-httpd (because it
always waits for HTTPD, as I understand things). It turns out we're
not really using this flag for anything, so we should just remove it.
* Scripts/webkitpy/tool/steps/runtests.py:
* Scripts/webkitpy/tool/steps/steps_unittest.py:
2011-05-01 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
The bots should learn from expected failures without having to retry
https://bugs.webkit.org/show_bug.cgi?id=59279
Rather than re-trying the whole patch, we can just check whether we now
expect the failures we see with the patch applied, saving a lot of time.
This patch also teaches the CommitQueue how to reject patches even
when the tree is red using the new unexpected_failures() method on
ExpectedFailures.
Finally in testing we found a bug in ExpectedFailures where it would
"trust" results which had many many failures, due to using an
equal sign instead of a less-than sign.
* Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py
* Scripts/webkitpy/tool/bot/expectedfailures.py
* Scripts/webkitpy/tool/bot/expectedfailures_unittest.py
* Scripts/webkitpy/tool/bot/patchanalysistask.py
2011-05-02 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
workspace.py fails to import ScriptError
https://bugs.webkit.org/show_bug.cgi?id=59915
If the workspace fails to create a zip, webkitpy crashes instead of
handling the error as expected.
* Scripts/webkitpy/common/system/workspace.py:
* Scripts/webkitpy/common/system/workspace_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
2011-05-01 Adam Barth <abarth@webkit.org>
Remove unadvertantly added file.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2011-05-01 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
expectedfailures.py should report the unexpected failures when asked
https://bugs.webkit.org/show_bug.cgi?id=59916
This information is useful when the EWS runs tests.
* Scripts/webkitpy/tool/bot/expectedfailures.py:
* Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
2011-05-01 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
JSONTestResult needs to handle multiple results
https://bugs.webkit.org/show_bug.cgi?id=59269
This case occurs when a test is flaky because NRWT will run the test
twice and report both results as "actual."
* Scripts/webkitpy/common/net/resultsjsonparser.py:
* Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2011-05-01 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Move checkout/api.py to checkout/checkout.py
https://bugs.webkit.org/show_bug.cgi?id=59906
This is more consistent with Python style.
* Scripts/webkitpy/common/checkout/__init__.py:
* Scripts/webkitpy/common/checkout/scm_unittest.py:
* Scripts/webkitpy/common/host.py:
* Scripts/webkitpy/tool/main.py:
2011-05-01 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
scm.py should be split into many pieces
https://bugs.webkit.org/show_bug.cgi?id=59908
Forgot an AuthenticationError import.
* Scripts/webkitpy/common/checkout/scm/git.py:
* Scripts/webkitpy/common/checkout/scm/svn.py:
2011-05-01 Arno Renevier <arno@renevier.net>
Reviewed by Eric Seidel.
wrong exception raised in Lighttpd start method
https://bugs.webkit.org/show_bug.cgi?id=59283
raise standard Exception instead of non existent
google.httpd_utils.HttpdNotStarted
* Scripts/webkitpy/layout_tests/port/http_server.py:
2011-05-01 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
scm.py should be split into many pieces
https://bugs.webkit.org/show_bug.cgi?id=59908
* Scripts/webkitpy/common/checkout/scm/__init__.py:
* Scripts/webkitpy/common/checkout/scm/commitmessage.py: Added.
* Scripts/webkitpy/common/checkout/scm/git.py: Added.
* Scripts/webkitpy/common/checkout/scm/scm.py:
* Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
* Scripts/webkitpy/common/checkout/scm/svn.py: Added.
2011-05-01 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Make most scm.py tests pass in preparation for splitting them up
https://bugs.webkit.org/show_bug.cgi?id=59904
There is one remaining failure:
webkitpy.common.checkout.scm.scm_unittest.GitTestWithMock.test_create_patch
which identifies a difference between the SVN and Git versions of the
SCM api. I'm not going to bother fixing that until we split out all
the tests.
* Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2011-05-01 Patrick Gansterer <paroga@webkit.org>
[WIN] Unreviewed buildfix after r85440.
* DumpRenderTree/win/UIDelegate.cpp:
(UIDelegate::UIDelegate): Removed accidental added adoptPtr.
2011-05-01 Patrick Gansterer <paroga@webkit.org>
[WIN] Unreviewed buildfix after r85434.
* DumpRenderTree/win/FrameLoadDelegate.cpp: Added missing header.
* DumpRenderTree/win/UIDelegate.cpp: Ditto.
2011-05-01 Patrick Gansterer <paroga@webkit.org>
[WIN] Unreviewed buildfix after r85434.
* DumpRenderTree/win/FrameLoadDelegate.cpp:
(FrameLoadDelegate::FrameLoadDelegate):
* DumpRenderTree/win/UIDelegate.cpp:
(DRTUndoManager::DRTUndoManager):
(UIDelegate::UIDelegate):
(UIDelegate::resetUndoManager):
2011-05-01 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Move scm.py into its own module in preparation for splitting it up
https://bugs.webkit.org/show_bug.cgi?id=59884
* Scripts/webkitpy/common/checkout/scm/__init__.py: Added.
* Scripts/webkitpy/common/checkout/scm/scm.py: Moved.
* Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Moved.
* Scripts/webkitpy/test/main.py:
2011-05-01 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
webkit-patch fails when not run from root
https://bugs.webkit.org/show_bug.cgi?id=59882
We need to set the CWD when running svn so that svn interprets file
paths relative to the root of the working copy.
* Scripts/webkitpy/common/checkout/scm.py:
2011-04-29 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
land-safely patches should show show their place in the commit-queue
https://bugs.webkit.org/show_bug.cgi?id=59853
This complicated the "should we show the bubbles or the submit-to-ews button"
logic substantially. :( But now we'll show both the commit-queue position
as well as the submit-to-ews button for land-safely (cq+ only) patches.
Previously we would only show the submit-to-ews button.
* QueueStatusServer/handlers/statusbubble.py:
* QueueStatusServer/templates/statusbubble.html:
2011-04-29 Naoki Takano <takano.naoki@gmail.com>
Reviewed by Adam Barth.
Add my name to contributers.
https://bugs.webkit.org/show_bug.cgi?id=59854
* Scripts/webkitpy/common/config/committers.py:
2011-04-29 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Csaba Osztrogonác.
[Qt] Don't build TestNetscapePlugin if ENABLE_NETSCAPE_PLUGIN_API=0
https://bugs.webkit.org/show_bug.cgi?id=59772
* Tools.pro:
2011-04-29 Stephanie Lewis <slewis@apple.com>
Rubber-stamped by Adam Roben.
Remove the production configuration guard since installs only happen in production anyway.
Works around a bug in Xcode
* DumpRenderTree/mac/PerlSupport/Makefile:
2011-04-29 Adam Roben <aroben@apple.com>
Remove a PHP setting that is now provided by httpd.conf
r85155 added this setting, but putting it in httpd.conf makes more sense (since the setting
isn't dependent on paths that could be different on each user's machine).
Reviewed by Sam Weinig.
* Scripts/webkitperl/httpd.pm:
(getDefaultConfigForTestDirectory): Removed log_errors setting.
2011-04-29 Jon Lee <jonlee@apple.com>
Unreviewed, Adding Jon Lee as a committer
* Scripts/webkitpy/common/config/committers.py:
2011-04-28 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
fix pretty_patch_available and wdiff_available
https://bugs.webkit.org/show_bug.cgi?id=59761
These methods rely on the check_* method being called first.
This would only happen when we actually tried to compute a diff,
but computing the diff happens on a different process.
Also fix setting _wdiff_avialable to True by default.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
2011-04-29 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
exclude skipped tests from the summarized results
https://bugs.webkit.org/show_bug.cgi?id=59813
If we find a need fo skipped tests later, we should either create
a skipped.json file or expose this some other way. As it is, the data
is unused and is hugely bloating the size of full_results.json.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-04-29 Adam Roben <aroben@apple.com>
Make DumpRenderTreeSupport get installed when the default configuration is built
The default configuration is Production, but in that case the CONFIGURATION environment
variable is set to the empty string.
Reviewed by Sam Weinig.
* DumpRenderTree/mac/PerlSupport/Makefile: Set CONFIGURATION to Production if it isn't
already set so that later logic that depends on CONFIGURATION having that value will work.
2011-04-29 Dean Jackson <dino@apple.com>
Reviewed by Simon Fraser.
Add ENABLE macro for WebKitAnimation
https://bugs.webkit.org/show_bug.cgi?id=59729
Add new feature to toggle WebKit Animation API.
* Scripts/build-webkit:
2011-04-29 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
Make production the default build type for WebKitTestRunner, matching
all other projects.
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2011-04-29 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix, add new directory to the build.
* waf/build/settings.py:
2011-04-29 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Implement UI client in minibrowser
https://bugs.webkit.org/show_bug.cgi?id=58416
It also allows to open more than one minibrowser window from the
command line.
* MiniBrowser/gtk/BrowserWindow.c:
(browserWindowFinalize):
(browser_window_init):
(browserWindowConstructed):
(createNewPage):
(showPage):
(closePage):
(runJavaScriptAlert):
(browserWindowUIClientInit):
* MiniBrowser/gtk/main.c:
(loadURI):
(main):
2011-04-29 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Implement page loader client in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=58024
A new class BrowserWindow has been added to implement the page
loader client. It makes supporting multiple windows easier. Window
title and url entry are now updated during the loading process.
* MiniBrowser/gtk/BrowserWindow.c: Added.
(activateUriEntryCallback):
(goBackCallback):
(goForwardCallback):
(browserWindowFinalize):
(browserWindowGetProperty):
(browserWindowSetProperty):
(browser_window_init):
(browserWindowConstructed):
(browser_window_class_init):
(browserWindowUpdateTitle):
(browserWindowSetTitle):
(browserWindowSetLoadProgress):
(browserWindowUpdateURL):
(didStartProvisionalLoadForFrame):
(didReceiveServerRedirectForProvisionalLoadForFrame):
(didFailProvisionalLoadWithErrorForFrame):
(didCommitLoadForFrame):
(didFinishDocumentLoadForFrame):
(didFinishLoadForFrame):
(didFailLoadWithErrorForFrame):
(didReceiveTitleForFrame):
(didFirstLayoutForFrame):
(didFirstVisuallyNonEmptyLayoutForFrame):
(didRemoveFrameFromHierarchy):
(didStartProgress):
(didChangeProgress):
(didFinishProgress):
(didBecomeUnresponsive):
(didBecomeResponsive):
(browserWindowLoaderClientInit):
(browser_window_new):
(browser_window_get_view):
* MiniBrowser/gtk/BrowserWindow.h: Added.
* MiniBrowser/gtk/GNUmakefile.am:
* MiniBrowser/gtk/main.c:
(createWindow):
2011-04-28 Stephanie Lewis <slewis@apple.com>
Reviewed by Mark Rowe.
http://bugs.webkit.org/show_bug.cgi?id=59743
<rdar://problem/9040413> Add an install target for production builds.
* DumpRenderTree/mac/PerlSupport/Makefile:
2011-04-28 Adam Roben <aroben@apple.com>
Stop and restart ATSServer before each test run on Leopard
This is to work around an ATSServer leak that eventually causes tests
to fail.
Fixes <http://webkit.org/b/51807> <rdar://problem/9273894>
fast/blockflow/broken-ideograph-[small-caps|font].html sometimes fails on Leopard buildbots
(due to ATSServer leak)
Reviewed by Dan Bernstein.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(StopATSServer):
(StartATSServer):
Added these new commands to stop and start ATSServer.
(TestFactory.__init__): Stop and start ATSServer on Leopard before running tests.
2011-04-28 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
add Leopard NRWT bot to the scheduler
https://bugs.webkit.org/show_bug.cgi?id=59766
* BuildSlaveSupport/build.webkit.org-config/config.json:
2011-04-28 Sam Weinig <sam@webkit.org>
Fix WebKitTestRunner build.
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
The Derived Sources target needs a product name.
2011-04-28 Sam Weinig <sam@webkit.org>
Reviewed by Mark Rowe.
Add install support for WebKitTestRunner.
<rdar://problem/9349055>
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
Add missing CompilerVersion.xcconfig.
* WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
* WebKitTestRunner/Configurations/WebKitTestRunner.xcconfig:
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
Add install path and tell it not to skip installation.
Add missing CompilerVersion.xcconfig.
2011-04-28 David Levin <levin@chromium.org>
Reviewed by Adam Barth.
Remove IMAGE_RESIZER related code.
https://bugs.webkit.org/show_bug.cgi?id=59735
* Scripts/build-webkit:
2011-04-28 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
Make reading skipped files from additional platform
directories work in WebKit2 mode.
* Scripts/old-run-webkit-tests:
(buildPlatformTestHierarchy): Simplify logic to always use the first
and last platforms in the build hierarchy as the test hierarchy, and to
insert the WebKit2 platform in the middle if it exists. Yum.
2011-04-28 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Remove WML
https://bugs.webkit.org/show_bug.cgi?id=59678
Remove build system integration and test harness integration for WML.
* Scripts/build-webkit:
* Scripts/old-run-webkit-tests:
* Scripts/webkitperl/features.pm:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-04-28 Ojan Vafai <ojan@chromium.org>
Reviewed by Mihai Parparita.
use a hierarchical data structure to store results json
https://bugs.webkit.org/show_bug.cgi?id=59736
Instead of a map from test name to results dict, we now store a hierarchical
data structure.
Old: { foo/bar/baz.html: results_dict, foo/bar/baz2.html: results_dict }
New:
{ foo: {
bar: {
baz.html: results_dict,
baz2.html: results_dict } } }
This cut ~52% of the size of times_ms.json. I expect it will cut
30-40% of full_results.json and unexpected_results.json.
* Scripts/webkitpy/common/net/resultsjsonparser.py:
* Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
Using os.sep was just wrong since relative_test_filename canonicalizes separators.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/tool/commands/rebaselineserver.py:
Instead of plumbing through the change to all of the rebaselineserver
python and JS, just convert to the old format after reading in the json.
2011-04-28 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
EWS Bubbles should show commit-queue position when patch is in the commit-queue
https://bugs.webkit.org/show_bug.cgi?id=59744
Add target="_top" to the bubble links. This was
forgotten when I moved from using an onclick to a normal link.
* QueueStatusServer/templates/statusbubble.html:
2011-04-28 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
EWS bubbles should show "submit to ews" button if they've not been processed
https://bugs.webkit.org/show_bug.cgi?id=59751
* QueueStatusServer/handlers/statusbubble.py:
* QueueStatusServer/handlers/submittoews.py:
* QueueStatusServer/templates/statusbubble.html:
* QueueStatusServer/templates/submittoews.html:
2011-04-28 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
EWS Bubbles should show commit-queue position when patch is in the commit-queue
https://bugs.webkit.org/show_bug.cgi?id=59744
This was actually just a 2 line change to statusbubble.py.
I also cleaned up statusbubble.html a little while I was there
(the links are now actual <a> tags and have hover).
* QueueStatusServer/handlers/statusbubble.py:
* QueueStatusServer/templates/statusbubble.html:
2011-04-28 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
add a mac leopard release NRWT bot
https://bugs.webkit.org/show_bug.cgi?id=59733
* BuildSlaveSupport/build.webkit.org-config/config.json:
2011-04-28 Ojan Vafai <ojan@chromium.org>
Avoid crashing when running new-run-webkit-tests outside of
version control.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-04-28 Xan Lopez <xlopez@igalia.com>
Unreviewed build fix.
* GNUmakefile.am:
2011-04-28 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
fix python printing test
https://bugs.webkit.org/show_bug.cgi?id=59719
Now that we run scm commands in summarize_results, --verbose prints
debug logging to stderr.
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2011-04-28 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
link to tests at a specific revision in the results.html file
https://bugs.webkit.org/show_bug.cgi?id=59711
* Scripts/webkitpy/common/checkout/scm.py:
Add a method to get the head svn revision.
* Scripts/webkitpy/common/checkout/scm_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
Include the svn revision in the results json.
2011-04-28 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Sam Weinig.
[GTK] Remove WKURLCreateWithURL()
https://bugs.webkit.org/show_bug.cgi?id=59691
Use WKURLCreateWithUTF8CString() instead of WKURLCreateWithURL().
* MiniBrowser/gtk/main.c:
(activateUriEntryCallback):
(main):
2011-04-28 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] WKView should be a GtkWidget
https://bugs.webkit.org/show_bug.cgi?id=59603
Don't use WKViewGetWindow(), use WKView as a GtkWidget instead.
* MiniBrowser/gtk/main.c:
(createWindow):
(main):
2011-04-28 Gabor Rapcsanyi <rgabor@webkit.org>
Reviewed by Csaba Osztrogonác.
[Qt] Add ARM specific Skipped list
https://bugs.webkit.org/show_bug.cgi?id=59680
* Scripts/old-run-webkit-tests:
* Scripts/webkitdirs.pm:
2011-04-27 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
error_log.txt should include PHP errors
https://bugs.webkit.org/show_bug.cgi?id=59643
Turns out this was really easy after I finally learned
enough about PHP to do it...
I'm adding this in hopes of better tracking down
our random http failures.
* Scripts/webkitperl/httpd.pm:
2011-04-27 David Levin <levin@chromium.org>
Reviewed by Adam Barth.
check-webkit-style should flags spacing errors for braces on the same line.
https://bugs.webkit.org/show_bug.cgi?id=59665
* Scripts/webkitpy/style/checkers/cpp.py: Added the check.
* Scripts/webkitpy/style/checkers/cpp_unittest.py: Added some tests for the check
and fixed some test cases to not trigger this error since it isn't what they are verifying.
2011-04-27 David Levin <levin@chromium.org>
Reviewed by Adam Barth.
check-webkit-style error for spaces after periods in a comment has a bad error
message and warns about end of line spaces.
https://bugs.webkit.org/show_bug.cgi?id=59663
* Scripts/webkitpy/style/checkers/cpp.py: Fix both issues.
* Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests for spaces at the end of line
and changed the error messages.
2011-04-27 Ojan Vafai <ojan@chromium.org>
remove assertion now that platform/mac-tiger no longer exists
https://bugs.webkit.org/show_bug.cgi?id=59660
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2011-04-27 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
remove test that didn't make sense
https://bugs.webkit.org/show_bug.cgi?id=59658
It was artifically saying the test wasn't expected to pass,
but the expectations string was empty. The case of unexpected
passes with a real expectations string is tested later in the
same test.
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2011-04-27 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
WebKit2 needs to notify the client when a plugin fails to initialize
https://bugs.webkit.org/show_bug.cgi?id=59657
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::initialize):
Add stubs.
2011-04-26 Sam Weinig <sam@webkit.org>
Reviewed by David Levin.
Add a few more missing options to new-run-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=59567
Add support for:
--threaded
--gc-between-tests
--complex-text
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-27 Mark Rowe <mrowe@apple.com>
Build fix.
Fix an OwnPtr-related issue that was missed before the strictness was turned on.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::runTest):
2011-04-27 Ojan Vafai <ojan@chromium.org>
Forgot to upload the new times_ms.json file after splitting
it out from full_results.json.
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-04-27 Ojan Vafai <ojan@chromium.org>
Fix snafu in generating the JSON for the chromium tests.
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2011-04-27 Ojan Vafai <ojan@chromium.org>
Fix bad merge. This broke the chromium tests generating JSON files.
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2011-04-27 Mark Rowe <mrowe@apple.com>
Rubber-stamped by Geoff Garen.
Don't look for results in a location that no longer exists.
* Scripts/old-run-webkit-tests:
2011-04-27 Adrienne Walker <enne@google.com>
Unreviewed, add Chrome GPU folks to contributors that I'm tired of manually CCing
* Scripts/webkitpy/common/config/committers.py:
2011-04-22 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
only include failures in full_results.json
https://bugs.webkit.org/show_bug.cgi?id=59256
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-04-22 Ojan Vafai <ojan@chromium.org>
Reviewed by Adam Roben.
move times into their own json file
https://bugs.webkit.org/show_bug.cgi?id=59239
The test runtimes are the only case where we need an entry for
every test that was run, including passing tests. Move them into
their own json file so we can prune the other json files to only
include failing tests.
Use a trie-like data structure to hold the tests to avoid including
full test paths. This cuts 52% of the filesize. The end result is ~700k
pre-gzip.
* Scripts/webkitpy/common/net/resultsjsonparser.py:
Exclude times from the parsed TestResults.
* Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/manager.py:
2011-04-27 Anders Carlsson <andersca@apple.com>
More casts and more clang++ build fixes.
* DumpRenderTree/mac/DumpRenderTree.mm:
(createWebViewAndOffscreenWindow):
* TestWebKitAPI/mac/PlatformWebViewMac.mm:
(TestWebKitAPI::PlatformWebView::PlatformWebView):
* WebKitTestRunner/mac/PlatformWebViewMac.mm:
(WTR::PlatformWebView::PlatformWebView):
2011-04-27 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
lazily autoinstall thirdparty python libraries
https://bugs.webkit.org/show_bug.cgi?id=55723
This helps by only having pywebsocket installed when the bots run. It
doesn't help much for devs since webkit-patch pulls in almost everything.
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/networktransaction.py:
* Scripts/webkitpy/common/net/networktransaction_unittest.py:
* Scripts/webkitpy/common/net/statusserver.py:
* Scripts/webkitpy/layout_tests/port/websocket_server.py:
* Scripts/webkitpy/thirdparty/__init__.py: Add an import hook so we can install based on
the import command rather than just installing everything.
* Scripts/webkitpy/thirdparty/__init___unittest.py: Added.
2011-04-27 Girish Ramakrishnan <girish@forwardbias.in>
Add my IRC nick.
* Scripts/webkitpy/common/config/committers.py:
2011-04-27 Eric Seidel <eric@webkit.org>
Reviewed by Mihai Parparita.
sherrifbot create-bug shouldn't assign bugs to webkit.review.bot
https://bugs.webkit.org/show_bug.cgi?id=59545
To do this, I needed a way to look up contributors by irc-name
(since anyone in #webkit who might use this command may not be a committer).
To lookup contributors, I had to make Contributor a real object.
Which led me to redesign parts of committers.py...
and finally fix one spot in changelog.py where we wanted to be
looking up contributors and not committers.
Overall a pretty simple fix, once you wade through the yak-hair.
This may not prevent *all* possible ways that bugs would get assigned
to webkit.review.bot. If we don't recognize the requester we will
go through the previous code path (which shouldn't change the assignee
on the bug from the default as far as I can tell).
* Scripts/webkitpy/common/checkout/changelog.py:
* Scripts/webkitpy/common/config/committers.py:
* Scripts/webkitpy/common/config/committers_unittest.py:
* Scripts/webkitpy/tool/bot/irc_command.py:
* Scripts/webkitpy/tool/bot/irc_command_unittest.py:
2011-04-27 Yi Shen <yi.4.shen@nokia.com>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Upstream Symbian platform plugin
https://bugs.webkit.org/show_bug.cgi?id=58435
Exempting directory WebKit/qt/symbian/platformplugin from style guide.
* Scripts/webkitpy/style/checker.py:
* Scripts/webkitpy/style/checker_unittest.py:
2011-04-27 Jeff Miller <jeffm@apple.com>
Correct my IRC nickname.
* Scripts/webkitpy/common/config/committers.py:
2011-04-27 Lucas De Marchi <lucas.demarchi@profusion.mobi>
Reviewed by Darin Adler.
Change the email used on bugzilla
https://bugs.webkit.org/show_bug.cgi?id=59582
* Scripts/webkitpy/common/config/committers.py:
2011-04-27 James Kozianski <koz@chromium.org>
Reviewed by Ojan Vafai.
Extract map from port to builder name.
https://bugs.webkit.org/show_bug.cgi?id=58301
* Scripts/webkitpy/layout_tests/port/__init__.py:
* Scripts/webkitpy/layout_tests/port/builders.py: Added.
* Scripts/webkitpy/layout_tests/port/builders_unittest.py: Added.
* Scripts/webkitpy/layout_tests/port/factory.py:
2011-04-27 Ryuan Choi <ryuan.choi@samsung.com>
Reviewed by Martin Robinson.
[GTK] Add proxy support to GtkLauncher
https://bugs.webkit.org/show_bug.cgi?id=58852
Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
* GNUmakefile.am:
* GtkLauncher/main.c:
(main):
2011-04-27 Annie Sullivan <sullivan@chromium.org>
Reviewed by David Levin.
Adding myself as a Chromium contributor to committers.py.
https://bugs.webkit.org/show_bug.cgi?id=59576
* Scripts/webkitpy/common/config/committers.py:
2011-04-26 David Grogan <dgrogan@chromium.org>
Reviewed by David Levin.
Adding myself to Chromium contributors list in committers.py
https://bugs.webkit.org/show_bug.cgi?id=59565
* Scripts/webkitpy/common/config/committers.py:
2011-04-26 Chang Shu <cshu@webkit.org>
Reviewed by Laszlo Gombos.
WebKitTestRunner needs layoutTestController.addOriginAccessWhitelistEntry
https://bugs.webkit.org/show_bug.cgi?id=42544
Also implemented removeOriginAccessWhitelistEntry and resetOriginAccessWhitelists.
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::addOriginAccessWhitelistEntry):
(WTR::LayoutTestController::removeOriginAccessWhitelistEntry):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-26 Chris Fleizach <cfleizach@apple.com>
Reviewed by Adele Peterson.
AX: Add more text marker routines to DRT
https://bugs.webkit.org/show_bug.cgi?id=58949
Add a few more methods to help test the text marker system, including nextMarker, previousMarker and
stringForMarkerRange.
* DumpRenderTree/AccessibilityUIElement.cpp:
(nextTextMarkerCallback):
(previousTextMarkerCallback):
(stringForTextMarkerRangeCallback):
(AccessibilityUIElement::previousTextMarker):
(AccessibilityUIElement::nextTextMarker):
(AccessibilityUIElement::stringForTextMarkerRange):
(AccessibilityUIElement::getJSClass):
* DumpRenderTree/AccessibilityUIElement.h:
* DumpRenderTree/mac/AccessibilityUIElementMac.mm:
(AccessibilityUIElement::previousTextMarker):
(AccessibilityUIElement::nextTextMarker):
(AccessibilityUIElement::stringForTextMarkerRange):
2011-04-26 Dan Bernstein <mitz@apple.com>
Build fix.
* DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
* MiniBrowser/Configurations/CompilerVersion.xcconfig:
* WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
2011-04-26 Kevin Ollivier <kevino@theolliviers.com>
Rubberstamped by Eric Seidel.
Enable waf to be used to build other ports
https://bugs.webkit.org/show_bug.cgi?id=58213
* Scripts/build-webkit:
* Scripts/webkitdirs.pm:
* waf: Added.
* waf/build: Added.
* waf/build/build_utils.py: Added.
* waf/build/settings.py: Added.
* waf/build/waf_extensions.py: Added.
* waf/build/wxpresets.py: Added.
2011-04-26 Sam Weinig <sam@webkit.org>
Reviewed by David Hyatt.
Remove Datagrid from the tree
https://bugs.webkit.org/show_bug.cgi?id=59543
* Scripts/build-webkit:
Remove feature.
* iExploder/iexploder-1.3.2/htdocs/htmltags.in:
* iExploder/iexploder-1.7.2/src/html-tags/webkit:
Remove from fuzzer.
2011-04-26 Brent Fulgham <bfulgham@webkit.org>
Reviewed by Adam Roben.
Update WinLauncher with command-line arguments to allow the
transparency to be tested. Use --transparent to create a
transparent view. Use --desktop to have the view fill the
desktop area of the screen.
https://bugs.webkit.org/show_bug.cgi?id=58300
* WinLauncher/WinLauncher.cpp:
(usesLayeredWebView): New method to indicate that the
program is running with layered windows (alpha-blended
transparent windows).
(shouldUseFullDesktop): New method to indicate that the
program is running across the full desktop (less any
task bars.)
(resizeSubViews): Layered window mode doesn't have any
winapi child windows, so this routine is exited early
for that case.
(subclassForLayeredWindow): Subclass the window so we can
override its defaut message loop.
(computeFullDesktopFrame): Convenience function to get the
desktop region and update the frame dimensions appropriately.
(_tWinMain): Updated to accept the new '--transparent' and
'--desktop' arguments.
(InitInstance): Modified to exit early for the transparent
background case. WebKit will construct our primary window (with
appropriate settings) so we don't need to create one ourselves.
(WndProc):
(TransparentWndProc): New message loop for transparent windows,
since they need special handling to support dragging.
2011-04-26 Alexandre MAzari <amazari@igalia.com>
Reviewed by Xan Lopez.
[Gtk] Implement LayoutTestController::setPluginsEnabled
https://bugs.webkit.org/show_bug.cgi?id=59184
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::setPluginsEnabled):
2011-04-26 Chang Shu <cshu@webkit.org>
Reviewed by Laszlo Gombos.
WebKitTestRunner needs layoutTestController.setAllowFileAccessFromFileURLs
https://bugs.webkit.org/show_bug.cgi?id=57572
Remove the unnecessary code that updates preference.
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setAllowUniversalAccessFromFileURLs):
2011-04-26 Mario Sanchez Prada <msanchez@igalia.com>
Reviewed by Eric Seidel.
[GTK] accessibility/aria-combobox-expected.txt
https://bugs.webkit.org/show_bug.cgi?id=59399
Implement isExpanded(), as it's used in the layout test.
* DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
(AccessibilityUIElement::isExpanded): Implemented.
2011-04-26 Adrienne Walker <enne@google.com>
Reviewed by Kent Tamura.
[chromium] Fix memory leak in DRT due to OwnPtr<T*>
https://bugs.webkit.org/show_bug.cgi?id=59489
* DumpRenderTree/chromium/WebViewHost.h:
2011-04-26 Mario Sanchez Prada <msanchez@igalia.com>
Reviewed by Eric Seidel.
[GTK] Implement isChecked in DRT
https://bugs.webkit.org/show_bug.cgi?id=59385
* DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
(AccessibilityUIElement::isChecked): Implemented.
2011-04-26 Alejandro G. Castro <alex@igalia.com>
Unreviewed, rolling out r84949.
http://trac.webkit.org/changeset/84949
https://bugs.webkit.org/show_bug.cgi?id=58852
Broke GTK+ compilation
* GtkLauncher/main.c:
(main):
2011-04-26 Eric Seidel <eric@webkit.org>
Reviewed by Mihai Parparita.
sheriff-bot should have a create-bug command
https://bugs.webkit.org/show_bug.cgi?id=59491
This patch has some stderr unittest spew, but I will
fix it after my live demo.
* Scripts/webkitpy/tool/bot/irc_command.py:
* Scripts/webkitpy/tool/bot/irc_command_unittest.py:
* Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2011-04-26 Ryuan Choi <ryuan.choi@samsung.com>
Reviewed by Martin Robinson.
[GTK] Add proxy support to GtkLauncher
https://bugs.webkit.org/show_bug.cgi?id=58852
Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
* GtkLauncher/main.c:
(main):
2011-04-26 Dan Bernstein <mitz@apple.com>
Reviewed by Mark Rowe.
Choose the compiler based on the Xcode version for Snow Leopard debug builds.
* DumpRenderTree/mac/Configurations/Base.xcconfig:
* DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig: Added.
* MiniBrowser/Configurations/Base.xcconfig:
* MiniBrowser/Configurations/CompilerVersion.xcconfig: Added.
* TestWebKitAPI/Configurations/Base.xcconfig:
* TestWebKitAPI/Configurations/CompilerVersion.xcconfig: Added.
* WebKitTestRunner/Configurations/Base.xcconfig:
* WebKitTestRunner/Configurations/CompilerVersion.xcconfig: Added.
2011-04-26 Finnur Thorarinsson <finnur.webkit@gmail.com>
Reviewed by David Levin.
Adding myself as a Chromium contributors to committers.py.
https://bugs.webkit.org/show_bug.cgi?id=59444
* Scripts/webkitpy/common/config/committers.py:
2011-04-26 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
DumpRenderTree/chromium should play nice with strict OwnPtrs
https://bugs.webkit.org/show_bug.cgi?id=59458
All these changes are pretty straight forward.
* DumpRenderTree/chromium/CppBoundClass.cpp:
(GetterPropertyCallback::GetterPropertyCallback):
(CppBoundClass::bindGetterCallback):
* DumpRenderTree/chromium/CppBoundClass.h:
(CppBoundClass::bindProperty):
(CppBoundClass::bindFallbackCallback):
(CppBoundClass::bindFallbackMethod):
* DumpRenderTree/chromium/DRTDevToolsClient.cpp:
(DRTDevToolsClient::DRTDevToolsClient):
* DumpRenderTree/chromium/TestEventPrinter.cpp:
(TestEventPrinter::createDRTPrinter):
(TestEventPrinter::createTestShellPrinter):
* DumpRenderTree/chromium/TestEventPrinter.h:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
(TestShell::createMainWindow):
(TestShell::createDRTDevToolsClient):
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::geolocationClientMock):
(WebViewHost::speechInputController):
(WebViewHost::deviceOrientationClientMock):
(WebViewHost::reset):
(WebViewHost::navigate):
(WebViewHost::setPendingExtraData):
(WebViewHost::canvas):
* DumpRenderTree/chromium/WebViewHost.h:
2011-04-26 Adam Roben <aroben@apple.com>
Make DRT/win log loading errors the same way DRT/mac does
Fixes <http://webkit.org/b/59303> [Windows WebKit1 Tests]
http/tests/misc/will-send-request-returns-null-on-redirect.html failing since r84742
Reviewed by Alexey Proskuryakov.
* DumpRenderTree/win/ResourceLoadDelegate.cpp:
(ResourceLoadDelegate::descriptionSuitableForTestResult): Removed the workaround for
<rdar://problem/5064234>, which has long since been fixed, and then made this function match
-[NSError(DRTExtras) _drt_descriptionSuitableForTestResult] more closely.
2011-04-26 Dan Bernstein <mitz@apple.com>
REGRESSION (84876): build-webkit produces an error message
https://bugs.webkit.org/show_bug.cgi?id=59460
* Scripts/webkitdirs.pm:
2011-04-26 Mihai Parparita <mihaip@chromium.org>
Reviewed by Eric Seidel.
Fix timestamp comparison in check-for-weak-vtables-and-externals
https://bugs.webkit.org/show_bug.cgi?id=59416
We should only re-run the check-for-weak-vtables-and-externals script
if the executable is more recent than the last run.
* Scripts/check-for-weak-vtables-and-externals:
2011-04-26 Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
[Qt] Implement LayoutTestController::setAutofilled
https://bugs.webkit.org/show_bug.cgi?id=59439
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setAutofilled):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-04-25 Dan Bernstein <mitz@apple.com>
Reviewed by Mark Rowe.
Teach build-webkit where newer Xcode versions store the application-wide build settings user default.
* Scripts/webkitdirs.pm:
2011-04-25 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
Leaks Viewer throws "Cannot post cyclic structures" in Safari 5
https://bugs.webkit.org/show_bug.cgi?id=56090
Make LeaksViewer work with versions of browsers that don't support message
passing of cyclic structures.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParser.js:
(LeaksParser.this._worker.onmessage):
(LeaksParser):
(LeaksParser.prototype.addLeaksFile):
Use LeaksParserImpl directly if we don't support passing cyclic structures.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserImpl.js: Copied from BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js.
(LeaksParserImpl):
(LeaksParserImpl.prototype.addLeaksFile):
(LeaksParserImpl.prototype._incorporateLeaks):
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
(onmessage):
Factor out LeaksParser implementation into LeaksParserImpl.js.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
(workersSupportCyclicStructures):
Add function to detect if workers support passing cyclic structures.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html:
Add LeaksParserImpl.js include.
2011-04-25 Jeff Miller <jeffm@apple.com>
Reviewed by Timothy Hatcher.
Treat strings defined with UI_STRING_LOCALIZE_LATER_KEY() as debugging strings in extract-localizable-strings
https://bugs.webkit.org/show_bug.cgi?id=59373
* Scripts/extract-localizable-strings: Add UI_STRING_LOCALIZE_LATER_KEY() to isDebugMacro var.
2011-04-25 Andrew Scherkus <scherkus@chromium.org>
Reviewed by Dimitri Glazkov.
Adding a few Chromium contributors to committers.py.
https://bugs.webkit.org/show_bug.cgi?id=59327
* Scripts/webkitpy/common/config/committers.py:
2011-04-23 Dominic Cooney <dominicc@chromium.org>
Reviewed by Dimitri Glazkov.
[V8] Nodes in shadow DOM should not be GCed while their hosts are alive
https://bugs.webkit.org/show_bug.cgi?id=59284
Chromium DRT over-refcounted the NPObject wrappers it used to
marshal shadowRoot nodes back to tests (WebBindings::makeNode
births NPObjects with referenceCount == 1). This masked the GC bug
59284 in tests by making shadowRoot nodes live forever.
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::shadowRoot):
(LayoutTestController::ensureShadowRoot):
2011-04-23 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fixes. Add new methods to wx LayoutTestController,
and update the URL for downloading waf.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::ensureShadowRoot):
(LayoutTestController::removeShadowRoot):
* Scripts/webkitdirs.pm:
2011-04-22 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r84627.
http://trac.webkit.org/changeset/84627
https://bugs.webkit.org/show_bug.cgi?id=59271
It broke Symbian build (Requested by Ossy on #webkit).
* Scripts/webkitpy/style/checker.py:
* Scripts/webkitpy/style/checker_unittest.py:
2011-04-22 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
rebaseline-chromium-webkit-tests: clean up output
https://bugs.webkit.org/show_bug.cgi?id=59240
This change dramatically changes the output the tool provides
by default. The previous output (more or less) is still there
in -v / verbose / debug mode, but the default output is
way terser and more readable. It mostly says which files are
updated, which are duplicates, and which are not found in an
archive. Feedback desired :).
This change also adds a scm.exists() method to the SCM object
so that we can cleanly figure out what we're doing while
rebaselining.
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/common/checkout/scm.py:
* Scripts/webkitpy/common/checkout/scm_unittest.py:
2011-04-22 Ilya Sherman <isherman@chromium.org>
Reviewed by Alexey Proskuryakov.
Add Ilya Sherman (isherman@chromium.org) to the list of contributors
https://bugs.webkit.org/show_bug.cgi?id=59231
* Scripts/webkitpy/common/config/committers.py:
2011-04-22 Yuta Kitamura <yutak@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: Exception raised when stderr contains binary data
https://bugs.webkit.org/show_bug.cgi?id=59083
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
Use write_binary_file() instead of write_text_file().
2011-04-22 Chang Shu <cshu@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
WebKitTestRunner needs layoutTestController.setFrameFlatteningEnabled
https://bugs.webkit.org/show_bug.cgi?id=42536
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setFrameFlatteningEnabled):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-22 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
CommitQueue should remember the expected failures from patch to patch
https://bugs.webkit.org/show_bug.cgi?id=59253
Previously, we stored the set of expected failures on the task object,
which meant the commit-queue forgot about them after each patch. This
patch moves them to the CommitQueue object so they will have a longer
lifetime.
* Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
* Scripts/webkitpy/tool/bot/patchanalysistask.py:
* Scripts/webkitpy/tool/commands/queues.py:
2011-04-22 Sam Weinig <sam@webkit.org>
Reviewed by Brian Weinstein.
WebKit2: Implement JavaScript unresponsiveness timer/callbacks
https://bugs.webkit.org/show_bug.cgi?id=59248
Stub out shouldInterruptJavaScript callbacks.
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]):
* MiniBrowser/win/BrowserView.cpp:
(BrowserView::create):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage):
(WTR::TestController::initialize):
2011-04-22 Daniel Bates <dbates@webkit.org>
Reviewed by Csaba Osztrogonác.
webkit-patch land hangs if svn prompts for credentials
https://bugs.webkit.org/show_bug.cgi?id=31500
Make has_authorization_for_realm() return True only if there exists a credentials file
that contains either the word "password" or "passtype". We assume that these words don't
coincide with the actual credential data (e.g. a person's username is "password").
* Scripts/webkitpy/common/checkout/scm.py:
* Scripts/webkitpy/common/checkout/scm_unittest.py:
- Added test cases:
test_has_authorization_for_realm_using_credentials_with_passtype(),
test_has_authorization_for_realm_using_credentials_with_password(),
test_not_have_authorization_for_realm_with_credentials_missing_password_and_passtype()
- Renamed test_not_have_authorization_for_realm() to test_not_have_authorization_for_realm_when_missing_credentials_file()
to better describe what it's testing.
- Repurposed test_has_authorization_for_realm() to take realm and credential data to use
and return the result of calling has_authorization_for_realm() so that the caller can
assert the result; Renamed to _test_has_authorization_for_realm_using_credentials() to
better reflect its new purpose.
2011-04-22 Adam Barth <abarth@webkit.org>
Remove double-negative. (Sorry this was bugging me.)
* Scripts/webkitpy/tool/bot/commitqueuetask.py:
2011-04-22 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Factor LayoutTestResultsReader out of CommitQueue
https://bugs.webkit.org/show_bug.cgi?id=59244
This code will be shared with the EWS when they start running tests.
* Scripts/webkitpy/tool/bot/layouttestresultsreader.py: Added.
* Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py: Added.
* Scripts/webkitpy/tool/commands/queues.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
2011-04-22 Eric Seidel <eric@webkit.org>
Reviewed by Dimitri Glazkov.
webkit-patch rebaseline is broken
https://bugs.webkit.org/show_bug.cgi?id=59207
My recent changes to how LayoutTestResults.results_from_string
uncovered a bug in our use of urllib2.
We now are treating the return value from _fetch_results_html()
(which returns urllib2.openurl()) as a string. That's incorrect,
the return value is a file-like object (which auto-converts to a
string in most cases).
I've updated our urllib2-using code to correctly call .read() on the
resulting object before treating it like a string.
Unfortunately there is no easy way to test this mis-understanding
of the openurl() semantics.
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
2011-04-22 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Factor PatchAnalysis task out of CommitQueueTask
https://bugs.webkit.org/show_bug.cgi?id=59220
This is a first step towards teaching the EWS to run tests.
* Scripts/webkitpy/tool/bot/commitqueuetask.py:
* Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
* Scripts/webkitpy/tool/bot/patchanalysistask.py: Added.
2011-04-22 Philippe Normand <pnormand@igalia.com>
Reviewed by Martin Robinson.
[GTK] DRT: implement eventSender.scalePageBy
https://bugs.webkit.org/show_bug.cgi?id=59082
* DumpRenderTree/gtk/EventSender.cpp:
(scalePageByCallback): scalePageBy() support.
2011-04-21 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
LayoutTestResults should know how to handle NRWT json files
https://bugs.webkit.org/show_bug.cgi?id=59168
LayoutTestResults was originally created to be a parallel
to existing NRWT architecture for dealing with results files.
But then it turns out that NRWT has no such architecture. :(
So this patch writes the necessary code to handle reading
full_results.json and unexpected_results.json files from
NRWT layout-test-results directories.
LayoutTestResults has thus morphed from being ORWT-only to
supporting both ORWT and NRWT. It's possible at some future
point that other pieces of the NRWT architecture will learn
how to read JSON files in which case this can go away.
This is all done for making it possible for the commit-queue
to run the tests using NWRT and be able to understand the results
(for flaky test reporting, etc.)
* Scripts/webkitpy/common/net/layouttestresults.py:
* Scripts/webkitpy/common/net/layouttestresults_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
* Scripts/webkitpy/common/net/resultsjsonparser.py: Added.
* Scripts/webkitpy/common/net/resultsjsonparser_unittest.py: Added.
2011-04-22 Yi Shen <yi.4.shen@nokia.com>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Upstream Symbian platform plugin
https://bugs.webkit.org/show_bug.cgi?id=58435
Exempting directory WebKit/qt/symbian/platformplugin from style guide.
* Scripts/webkitpy/style/checker.py:
* Scripts/webkitpy/style/checker_unittest.py:
2011-04-22 Balazs Kelemen <kbalazs@webkit.org>
Reviewed by Csaba Osztrogonác.
[Qt][WK2] Qt should use the mac-wk2 skipped list
https://bugs.webkit.org/show_bug.cgi?id=58536
* Scripts/old-run-webkit-tests:
2011-04-21 Tony Chang <tony@chromium.org>
Reviewed by Mihai Parparita.
Remove find-mismatched-layout-test-results since it's not needed in a
checksum free world.
https://bugs.webkit.org/show_bug.cgi?id=59156
* Scripts/find-mismatched-layout-test-results: Removed.
2011-04-21 Brent Fulgham <bfulgham@webkit.org>
Reviewed by Martin Robinson.
Teach build-webkit and update-webkit that the WinCairo port does
not need the non-redistributable WebKitSupportLibraries.zip file.
Also append the _Cairo_CFLite label to the build target for the
default (no argument) case when building with '--wincairo'
https://bugs.webkit.org/show_bug.cgi?id=58801
* Scripts/build-webkit: Don't die if WebKitSupportLibraries.zip
doesn't exist.
* Scripts/webkitdirs.pm: Append '_Cairo_CFLite' to the default
'Release' build target when passed the '--wincairo' flag.
* Scripts/update-webkit: Don't die if WebKitSupportLibraries.zip
doesn't exist.
2011-04-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
new-run-webkit-tests: obsolete old code, part 5: rename message_broker2 to message_broker
https://bugs.webkit.org/show_bug.cgi?id=58758
Since worker_mixin is now only used by worker, there's no point
in having it split across two files.
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Added
* Scripts/webkitpy/layout_tests/layout_package/message_broker2.py: Removed.
* Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: * Added
* Scripts/webkitpy/layout_tests/layout_package/message_broker2_unittest.py: Removed.
2011-04-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: obsolete old code, part 4: merge worker_mixin into worker
https://bugs.webkit.org/show_bug.cgi?id=58756
Since worker_mixin is now only used by worker, there's no point
in having it split across two files.
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
* Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py: Removed.
2011-04-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: obsolete old code, part 3 - rename test_runner to manager
https://bugs.webkit.org/show_bug.cgi?id=58754
test_runner hasn't been a great name for some time, since all of
the test running is actually done in worker and
single_test_runner. Rename to manager to match the terminology
elsewhere.
* Scripts/webkitpy/layout_tests/layout_package/manager.py: Added.
* Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Added.
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py: Removed.
* Scripts/webkitpy/layout_tests/layout_package/test_runner_unittest.py: Removed.
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-04-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: obsolete old threading code, part 2
https://bugs.webkit.org/show_bug.cgi?id=58753
This merges the test_runner2 code back into test_runner since
there's only a single type of test runner now.
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner2.py: Removed.
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-04-21 Maciej Stachowiak <mjs@apple.com>
Reviewed by Adam Roben.
Add a feature define to allow <details> and <summary> to be disabled
https://bugs.webkit.org/show_bug.cgi?id=59118
<rdar://problem/9257045>
* Scripts/build-webkit:
* Tools.pro:
2011-04-21 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
webkit-patch upload should assign the bug to you
https://bugs.webkit.org/show_bug.cgi?id=33699
This patch only re-assigns the bug if it is currently unassigned. We
can be more aggressive in the future if people want us to be more
agressive.
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/tool/commands/upload_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
* Scripts/webkitpy/tool/steps/postdiff.py:
2011-04-21 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
move json_results.html over to LayoutTests/fast/harness
https://bugs.webkit.org/show_bug.cgi?id=59154
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-21 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
ValidateChangeLogs doesn't work on SVN
https://bugs.webkit.org/show_bug.cgi?id=59115
svn-create-patch lies about the diff in the working copy by moving
ChangeLog entries to the top of the diff. That's fine on most cases,
but causes problems for ValidateChangeLogs, which is trying validate
the where the ChangeLog entry appears.
I haven't added a test for this change because I couldn't figure out
how to write one. The issue is more of an integration issue, which
we're not really set up to test in our unit testing framework. If this
patch had worked around the output from svn-create-patch, then I could
have tested that we behave correctly on sample svn-create-patch output,
but, in this case, I've removed the dependency on svn-create-patch. I
could test that we behave correctly on "svn diff" output, but we
already have those tests.
* Scripts/webkitpy/tool/commands/download_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
* Scripts/webkitpy/tool/steps/validatechangelogs.py:
2011-04-21 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
old-run-webkit-tests/new-run-webkit-tests shouldn't read or write .checksum files
https://bugs.webkit.org/show_bug.cgi?id=58402
* Scripts/old-run-webkit-tests: Don't read or write .checksum files.
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Don't write .checksum files.
* Scripts/webkitpy/layout_tests/port/base.py: Don't read .checksum files.
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/port/test.py: Update passing pngs to include a checksum.
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-21 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
final tests for results.html file
https://bugs.webkit.org/show_bug.cgi?id=59150
This gets us close to 100% code coverage. Once this is
checked in, I'll move this file over to LayoutTests/fast/harness
and make it a proper layout test.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
2011-04-21 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
only expand visible rows
https://bugs.webkit.org/show_bug.cgi?id=59130
For the chromium port, where there are many expected failures,
expanding all of them is quite slow. Only expand the visible ones.
This makes the default case. Also, make the expanding async to
avoid totally killing the process.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
2011-04-21 Martin Robinson <mrobinson@igalia.com>
[GTK] DRT: implement eventSender.scalePageBy
https://bugs.webkit.org/show_bug.cgi?id=59082
Rolling out r84507 since it broke many GTK+ tests.
* DumpRenderTree/gtk/EventSender.cpp:
2011-04-21 Sam Weinig <sam@webkit.org>
Reviewed by Simon Fraser.
Add way to garbage collect between each test when running layout tests
https://bugs.webkit.org/show_bug.cgi?id=59126
* DumpRenderTree/mac/DumpRenderTree.mm:
(initializeGlobalsFromCommandLineOptions):
(runTest):
* Scripts/old-run-webkit-tests:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::didReceiveMessage):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::TestController):
(WTR::TestController::initialize):
(WTR::TestController::resetStateToConsistentValues):
* WebKitTestRunner/TestController.h:
2011-04-21 Ojan Vafai <ojan@chromium.org>
Reviewed by Mihai Parparita.
fix another innocuous JS error in expand/collapseAllExpectations
https://bugs.webkit.org/show_bug.cgi?id=59122
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
2011-04-20 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
put unexpected passes into their own table
https://bugs.webkit.org/show_bug.cgi?id=59016
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-20 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
initial set of tests for new results.html file
https://bugs.webkit.org/show_bug.cgi?id=59031
Also, fix the bugs (mostly innocuous JS errors) the tests found.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/json_results_test.js: Added.
2011-04-21 Philippe Normand <pnormand@igalia.com>
Reviewed by Martin Robinson.
[GTK] missing LayoutTestController::callShouldCloseOnWebView()
https://bugs.webkit.org/show_bug.cgi?id=59086
callShouldCloseOnWebView() implementation.
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::callShouldCloseOnWebView):
2011-04-21 Philippe Normand <pnormand@igalia.com>
Reviewed by Martin Robinson.
[GTK] DRT: implement eventSender.scalePageBy
https://bugs.webkit.org/show_bug.cgi?id=59082
* DumpRenderTree/gtk/EventSender.cpp:
(scalePageByCallback): scalePageBy() support.
2011-04-20 Dominic Cooney <dominicc@chromium.org>
Reviewed by Dimitri Glazkov.
layoutTestController can create and destroy shadow DOM
https://bugs.webkit.org/show_bug.cgi?id=59058
Chromium, Mac, GTK and Qt DRT; and WK2 test runner.
* DumpRenderTree/LayoutTestController.cpp:
(ensureShadowRootCallback):
(removeShadowRootCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::ensureShadowRoot):
(LayoutTestController::removeShadowRoot):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::ensureShadowRoot):
(LayoutTestController::removeShadowRoot):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::shadowRoot):
(LayoutTestController::ensureShadowRoot):
(LayoutTestController::removeShadowRoot):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::ensureShadowRoot):
(LayoutTestController::removeShadowRoot):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::ensureShadowRoot):
(LayoutTestController::removeShadowRoot):
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::shadowRoot):
(WTR::LayoutTestController::ensureShadowRoot):
(WTR::LayoutTestController::removeShadowRoot):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-20 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
new-run-webkit-tests: handle ctrl-c more cleanly
https://bugs.webkit.org/show_bug.cgi?id=58849
* Scripts/new-run-webkit-tests:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
2011-04-20 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r84431.
http://trac.webkit.org/changeset/84431
https://bugs.webkit.org/show_bug.cgi?id=59061
Breaks for folks who don't have editbugs (Requested by abarth
on #webkit).
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/tool/commands/upload_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
* Scripts/webkitpy/tool/steps/postdiff.py:
2011-04-20 Leo Yang <leo.yang@torchmobile.com.cn>
Unreviewed.
Adding myself to committers.py.
* Scripts/webkitpy/common/config/committers.py:
2011-04-20 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
webkit-patch upload should assign the bug to you
https://bugs.webkit.org/show_bug.cgi?id=33699
This patch only re-assigns the bug if it is currently unassigned. We
can be more aggressive in the future if people want us to be more
agressive.
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/tool/commands/upload_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
* Scripts/webkitpy/tool/steps/postdiff.py:
2011-04-20 Peter Gal <galpeter@inf.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
Modify builder.html template to allow easy access
to list more builds on the builder's page.
https://bugs.webkit.org/show_bug.cgi?id=59008
* BuildSlaveSupport/build.webkit.org-config/templates/builder.html:
2011-04-20 Peter Gal <galpeter@inf.u-szeged.hu>
Reviewed by Csaba Osztrogonác.
Add the default template for builder page.
https://bugs.webkit.org/show_bug.cgi?id=59008
* BuildSlaveSupport/build.webkit.org-config/templates/builder.html: Added.
2011-04-20 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
move page generation into a single function
https://bugs.webkit.org/show_bug.cgi?id=59027
Now executing the script doesn't create any state. We should then
be able to write tests for this by setting custom results data.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-20 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
move global state into a shared object
https://bugs.webkit.org/show_bug.cgi?id=59023
This makes the code more testable.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-20 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
move global state into a shared object
https://bugs.webkit.org/show_bug.cgi?id=59023
This makes the code more testable.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-20 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
add expandAll/collapseAll links
https://bugs.webkit.org/show_bug.cgi?id=59011
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-20 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fixes for wx 2.9.1, Cocoa and 64-bit builds.
* Scripts/webkitdirs.pm:
* wx/build/build_utils.py:
* wx/build/settings.py:
* wx/install-unix-extras:
* wx/packaging/build-mac-installer.py:
* wx/packaging/build-win-installer.py:
2011-04-20 Adam Roben <aroben@apple.com>
Replace old-run-webkit-tests's win-7 platform with win-7sp0
It was an error that we had a win-7 platform to begin with. Windows 7 was the most recent
version of Windows we cared about, so its results should just go in win. But now that
Windows 7 SP1 exists (and has different results for some tests), we need a place to store
results for Windows 7 without SP1 (and older versions of Windows).
Part of <http://webkit.org/b/56617> Several tests fail on Windows 7 SP1 due to font
differences from Windows 7 "SP0"
Reviewed by David Kilzer.
* Scripts/old-run-webkit-tests: Replaced win-7 with win-7sp0. Newer versions of Windows 7
will just use "win".
* Scripts/webkitdirs.pm:
(isWindows7SP0): Added. Replaces isWindows7, because that function had no remaining callers.
2011-04-20 Adam Roben <aroben@apple.com>
Teach webkitdirs::winVersion about build numbers
The build number isn't currently used, but soon will be.
As a bonus, winVersion now works on non-Cygwin variants of Perl!
Fixes <http://webkit.org/b/58991> Need a way to distinguish between Windows 7 and Windows 7
SP1
Reviewed by Anders Carlsson.
* Scripts/webkitdirs.pm:
(isAnyWindows): Added. Returns true if we're running on any Windows-ish Perl.
(determineWinVersion): Use `cmd /c ver` to get the version number, including the build
number. Parse the version number into a hash, similar to how osXVersion() works.
(isWindows7):
(isWindowsVista):
(isWindowsXP):
Updated for changes to winVersion().
2011-04-20 W. James MacLean <wjmaclean@chromium.org>
Reviewed by Kenneth Russell.
Add support for composite-to-texture to DumpRenderTree.
https://bugs.webkit.org/show_bug.cgi?id=58675
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
(TestShell::resetWebSettings):
* DumpRenderTree/chromium/TestShell.h:
(TestShell::setCompositeToTexture):
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::reset):
(WebPreferences::applyTo):
* DumpRenderTree/chromium/WebPreferences.h:
2011-04-19 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
The commit-queue is confused when non-layout tests fail
https://bugs.webkit.org/show_bug.cgi?id=58955
As seen in http://queues.webkit.org/results/8474435
It fails to create an archive of the layout test results
and then throws an exception.
I've fixed this by making it catch the exception as well
as made it so it never calls that path in the common case.
I've updated MockFileSystem to understand rmtree's affect on
directories as well as files.
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/common/system/workspace.py:
* Scripts/webkitpy/tool/commands/queues.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
2011-04-19 Ojan Vafai <ojan@chromium.org>
Fix check for whether a failure is expected to deal with
flaky tests and FAIL expectations.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-19 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
don't show missing text results for tests that only dump image results
https://bugs.webkit.org/show_bug.cgi?id=58931
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-19 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
don't show failures table if there are only new tests
https://bugs.webkit.org/show_bug.cgi?id=58929
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-19 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
don't show failures table if there are only new tests
https://bugs.webkit.org/show_bug.cgi?id=58929
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-19 Ojan Vafai <ojan@chromium.org>
Fix unittest failures from http://trac.webkit.org/changeset/84294.
* Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner_unittest.py:
2011-04-18 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
switch new-run-webkit-tests to using the new results file
https://bugs.webkit.org/show_bug.cgi?id=58861
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
* Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-19 Renata Hodovan <reni@webkit.org>
Reviewed by Eric Seidel.
Move the alignment related macros from Vector.h to new Alignment.h.
https://bugs.webkit.org/show_bug.cgi?id=56000
Adding the new file to the forwarding headers of Tools.
* DumpRenderTree/ForwardingHeaders/wtf/Alignment.h: Added.
2011-04-19 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix for Win / wx 2.9 wxPython installation.
* wx/packaging/build-win-installer.py:
* wx/packaging/wxWebKitInstaller.iss.in:
2011-04-19 Adam Roben <aroben@apple.com>
Skip another accelerated-compositing-sensitive test when accelerated compositing is disabled
* Scripts/old-run-webkit-tests: Skip media/controls-without-preload.html on Windows when
accelerated compositing is disabled.
2011-04-19 Ojan Vafai <ojan@chromium.org>
Remove accidentally committed debug code.
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-19 Brent Fulgham <bfulgham@webkit.org>
Unreviewed build correction.
Avoid duplicate stub implementation of test routine
for WinCairo build.
* WebKitTestRunner/win/TestInvocationWin.cpp: Exclude
WinCairo from using this stub.
2011-04-18 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
make results file work with audio and reftests
https://bugs.webkit.org/show_bug.cgi?id=58860
Also fix bug with timeout tests and store a bit in the JSON
for new image tests instead of loading the image result to check if it's there.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-19 Jer Noble <jer.noble@apple.com>
Reviewed by Adam Roben.
REGRESSION (r84206): 50 tests failing on SnowLeopard Intel Release (WebKit2 Tests) due to extra "supportsFullScreen() == true" output
https://bugs.webkit.org/show_bug.cgi?id=58890
Add a layoutTestController flag which controls when full screen callbacks are written to stdout.
Defaults to false.
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpFullScreenCallbacks().
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::supportsFullScreen): Only write to stdout if shouldDumpFullScreenCallbacks is set.
(WTR::InjectedBundlePage::enterFullScreenForElement): Ditto.
(WTR::InjectedBundlePage::exitFullScreenForElement): Ditto.
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::LayoutTestController): Initialize new ivar.
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
(WTR::LayoutTestController::dumpFullScreenCallbacks): Added.
(WTR::LayoutTestController::shouldDumpFullScreenCallbacks): Added.
2011-04-19 Kristóf Kosztyó <Kosztyo.Kristof@stud.u-szeged.hu>
Reviewed by Andreas Kling.
[Qt] fast/xmlhttprequest/xmlhttprequest-nonexistent-file.html fails (DRT sideeffect)
https://bugs.webkit.org/show_bug.cgi?id=35086
fast/xmlhttprequest/xmlhttprequest-no-file-access.html caused
a DRT sideeffect problem, because the resetSettings didn't set
the default value of LocalContentCanAccessFileUrls.
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::WebPage::resetSettings):
2011-04-18 Zan Dobersek <zandobersek@gmail.com>
Reviewed by Martin Robinson.
[GTK] Need support for dumping focus rectangles in pixel results
https://bugs.webkit.org/show_bug.cgi?id=53647
Dump the selection rect if it has been requested.
* DumpRenderTree/gtk/PixelDumpSupportGtk.cpp:
(createBitmapContextFromWebView):
2011-04-18 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
test-webkitpy should run the ActiveWorkItems tests
https://bugs.webkit.org/show_bug.cgi?id=58859
Also make the unit tests pass. :)
* QueueStatusServer/model/activeworkitems.py:
* QueueStatusServer/model/queue.py:
* QueueStatusServer/model/workitems.py:
2011-04-18 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
queues.webkit.org should show trailing-days pass counts for queues
https://bugs.webkit.org/show_bug.cgi?id=58812
This lets us see if each bot is pulling its own weight or not.
Before I wrote this patch our suspicious was that one of the
cq bots was doing most of the work. Turns out they're actually
pretty even in their recent pass counts.
* QueueStatusServer/handlers/queuestatus.py:
* QueueStatusServer/index.yaml:
* QueueStatusServer/templates/queuestatus.html:
2011-04-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
new-run-webkit-tests: obsolete old threading code, part 1
https://bugs.webkit.org/show_bug.cgi?id=58752
This patch removes support for the 'old-inline' and
'old-threads' worker models, and deletes the corresponding code.
Now that the new code is stable, this stuff is no longer
necessary.
* Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py: Removed.
* Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
* Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
new-run-webkit-tests: add an --additional-drt-flag option
https://bugs.webkit.org/show_bug.cgi?id=58680
NRWT has a bunch of command line flags that exist to pass
custom flags to DRT, especially on chromium. It would be nice
if there was a generic mechanism to pass through flags so we
didn't have all the custom ones.
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-04-05 Jer Noble <jer.noble@apple.com>
Reviewed by Sam Weinig.
WebKit2: WKTR should support WebKit2 full screen APIs
https://bugs.webkit.org/show_bug.cgi?id=56318
Respond to enterFullScreenForElement() and exitFullScreenForElement(), allowing
WKTR to test the LayoutTest/fullscreen/ tests.
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage): Add support for WKBundlePageFullScreenClient.
(WTR::InjectedBundlePage::supportsFullScreen): Added.
(WTR::InjectedBundlePage::enterFullScreenForElement): Added.
(WTR::InjectedBundlePage::exitFullScreenForElement): Added.
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::resetStateToConsistentValues): Enable the full screen preference.
2011-04-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: remove chromium-specific relpath hack for --results-directory
https://bugs.webkit.org/show_bug.cgi?id=58388
* Scripts/webkitpy/layout_tests/port/chromium.py:
2011-04-18 Evan Martin <evan@chromium.org>
Reviewed by Eric Seidel.
[chromium] expose title direction to webkit client
https://bugs.webkit.org/show_bug.cgi?id=58823
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::reset):
* DumpRenderTree/chromium/LayoutTestController.h:
(LayoutTestController::setTitleTextDirection):
Add member and setter to let us stash the title direction at load time.
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::didReceiveTitle):
* DumpRenderTree/chromium/WebViewHost.h:
Adapt to new API; push the title into layoutTestController when we
receive it.
2011-04-18 Dirk Pranke <dpranke@chromium.org>
Unreviewed.
Re-land r84112. It was rolled out in r84127 but it didn't need
to be. Only r84111 actually broke things.
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: don't run more processes than necessary
https://bugs.webkit.org/show_bug.cgi?id=58751
Previously NRWT would start up 1 process for each processor on
the machine by default, which was a waste if we were only
running a few shards of tests.
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-18 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
include links to actual.png files for new pixel tests
https://bugs.webkit.org/show_bug.cgi?id=58831
Load image elements to confirm whether the files exist.
Image elements will load without being appended to the DOM.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-18 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
add a zoomed view for pixel results to the new results html file
https://bugs.webkit.org/show_bug.cgi?id=58827
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-18 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
add labels to results html page
https://bugs.webkit.org/show_bug.cgi?id=58803
-fix iframe size to be the size of the png results
-fix a bunch of aesthetic nits
-fix some incorrectly styled code
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-18 Chang Shu <cshu@webkit.org>
Reviewed by Eric Seidel.
WebKitTestRunner needs layoutTestController.pathToLocalResource
https://bugs.webkit.org/show_bug.cgi?id=42541
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
* WebKitTestRunner/InjectedBundle/mac/LayoutTestControllerMac.mm:
(WTR::LayoutTestController::pathToLocalResource):
* WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
(WTR::LayoutTestController::pathToLocalResource):
* WebKitTestRunner/InjectedBundle/win/LayoutTestControllerWin.cpp:
(WTR::LayoutTestController::pathToLocalResource):
2011-04-18 Timothy Hatcher <timothy@apple.com>
Make update-webkit-localizable-strings put WebKit/win strings in WebCore
now that all localized strings in WebKit/win use WEB_UI_STRING.
https://webkit.org/b/58747
Reviewed by Dan Bernstein.
* Scripts/update-webkit-localizable-strings: Remove the code that updated WebKit/win differently.
2011-04-18 Patrick Gansterer <paroga@webkit.org>
Reviewed by Adam Roben.
[WIN] Use WCHAR instead of TCHAR
https://bugs.webkit.org/show_bug.cgi?id=58755
We always use the UNICODE versions of windows functions, so
the usage of TCHAR makes no sense and mixing them is bad style.
* WinLauncher/PrintWebUIDelegate.cpp:
Also fix various style issues (including indentation and removal of ugly C-casts).
* WinLauncher/WinLauncher.cpp:
2011-04-18 Dominic Cooney <dominicc@chromium.org>
Reviewed by Andreas Kling.
Add layoutTestController.shadowRoot to Qt DRT.
https://bugs.webkit.org/show_bug.cgi?id=58759
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::shadowRoot): Added.
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-04-18 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r84111.
http://trac.webkit.org/changeset/84111
https://bugs.webkit.org/show_bug.cgi?id=58771
Chromium linux layout tests are broken. (Requested by loislo2
on #webkit).
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
* Scripts/webkitpy/layout_tests/port/mock_drt.py:
2011-04-18 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r84112.
http://trac.webkit.org/changeset/84112
https://bugs.webkit.org/show_bug.cgi?id=58770
Chromium linux layout tests are broken. (Requested by loislo2
on #webkit).
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-17 Daniel Bates <dbates@webkit.org>
Reviewed by Eric Seidel and Adam Barth.
webkit-patch land hangs if svn prompts for credentials
https://bugs.webkit.org/show_bug.cgi?id=31500
Prompt for username and password when using git-svn and there aren't cached SVN credentials.
* Scripts/webkitpy/common/checkout/scm.py:
- Added mixin class SVNRepository and made both class SVN and Git inherit from it.
- Moved SVN.has_authorization_for_realm() to class SVNRepository and removed default value
for argument realm; modified call sites as needed.
- Modified AuthenticationError constructor to take optional prompt_for_password argument.
- Modified {SCM, SVN, Git}.commit_with_message() to take optional password argument.
- Modified Git._commit_on_branch() to take optional username and password argument.
- Modified Git.push_local_commits_to_server() to take optional username and password
argument and to call has_authorization_for_realm().
* Scripts/webkitpy/common/checkout/scm_unittest.py:
- Modified SVNTest.test_commit_without_authorization() to take dummy realm argument.
- Modified SVNTest.test_not_have_authorization_for_realm() to pass realm argument to
SVN.has_authorization_for_realm().
* Scripts/webkitpy/common/net/credentials.py:
- Modified Credentials.read_credentials() to call User.prompt_password() instead
of using getpass.getpass() directly.
* Scripts/webkitpy/common/system/user.py:
- Added User.prompt_password().
* Scripts/webkitpy/tool/steps/commit.py:
- Modified Commit.run() to prompt for a password if needed.
2011-04-17 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: save stderr into the layout-test-results dir
https://bugs.webkit.org/show_bug.cgi?id=58690
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-17 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: read stderr from chromium DRT separately
https://bugs.webkit.org/show_bug.cgi?id=58708
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
* Scripts/webkitpy/layout_tests/port/mock_drt.py:
2011-04-17 Patrick Gansterer <paroga@webkit.org>
Reviewed by Adam Barth.
Rename PLATFORM(CA) to USE(CA)
https://bugs.webkit.org/show_bug.cgi?id=58742
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetDefaultsToConsistentValues):
2011-04-17 Patrick Gansterer <paroga@webkit.org>
Reviewed by Adam Barth.
Rename PLATFORM(CG) to USE(CG)
https://bugs.webkit.org/show_bug.cgi?id=58729
* DumpRenderTree/PixelDumpSupport.cpp:
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::paintRect):
* DumpRenderTree/config.h:
* DumpRenderTree/win/PixelDumpSupportWin.cpp:
(createBitmapContextFromWebView):
* TestWebKitAPI/PlatformWebView.h:
* WebKitTestRunner/config.h:
* WebKitTestRunner/win/TestInvocationWin.cpp:
2011-04-16 Patrick Gansterer <paroga@webkit.org>
Reviewed by Eric Seidel.
Rename PLATFORM(CAIRO) to USE(CAIRO)
https://bugs.webkit.org/show_bug.cgi?id=55192
* DumpRenderTree/PixelDumpSupport.cpp:
* DumpRenderTree/config.h:
* DumpRenderTree/win/PixelDumpSupportWin.cpp:
(createBitmapContextFromWebView):
* WebKitTestRunner/config.h:
2011-04-16 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
show diffs+test inline in the results page
https://bugs.webkit.org/show_bug.cgi?id=58723
Put each row in it's own tbody and sort based on tbodies. That way
we can append the results to the tbody inline as another table row.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-16 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
handle new tests in the new results html file
https://bugs.webkit.org/show_bug.cgi?id=58715
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-15 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
old-run-webkit-tests: save the list of tests actually run to a file
https://bugs.webkit.org/show_bug.cgi?id=58692
* Scripts/old-run-webkit-tests:
2011-04-15 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
fix MockFileSystem.abspath() on windows
https://bugs.webkit.org/show_bug.cgi?id=58713
The existing code was borked on Windows if a path was passed
in using a drive letter (e.g., "d:\foo.txt"), and would go
into an infinite recursive loop. While win paths shouldn't
normally be passed into the mock filesystem, this just
bulletproofs things a bit (note that "/foo.txt" is considered
an absolute path on windows python as well.
* Scripts/webkitpy/common/system/filesystem_mock.py:
2011-04-15 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
show stderr files and httpd log files
https://bugs.webkit.org/show_bug.cgi?id=58714
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-15 Shishir Agrawal <shishir@chromium.org>
Reviewed by James Robinson.
Add a flag to guard Page Visibility API changes.
https://bugs.webkit.org/show_bug.cgi?id=58464
* Scripts/build-webkit:
2011-04-15 Chang Shu <cshu@webkit.org>
Reviewed by Alexey Proskuryakov.
When a message with url embedded is added to console, the "file:" scheme
and path should be stripped.
https://bugs.webkit.org/show_bug.cgi?id=58665
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::lastFileURLPathComponent):
(WTR::InjectedBundlePage::willAddMessageToConsole):
2011-04-15 Jeff Miller <jeffm@apple.com>
Reviewed by Sam Weinig.
Add takeFocus callback to WKPageUIClient
https://bugs.webkit.org/show_bug.cgi?id=58686
Add empty WKPageUIClient entries for the takeFocus callback in the MiniBrowser and WebKitTestRunner.
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]): Added empty takeFocus entry.
* MiniBrowser/win/BrowserView.cpp:
(BrowserView::create): Added empty takeFocus entry.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage): Added empty takeFocus entry.
(WTR::TestController::initialize): Added empty takeFocus entry.
2011-04-15 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
only show pretty-diff/wdiff if they were generated
https://bugs.webkit.org/show_bug.cgi?id=58682
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/port/base.py:
2011-04-15 Andreas Kling <kling@webkit.org>
Reviewed by Antonio Gomes.
[Qt] DRT: Default custom policy delegate decision should be Ignore.
If layoutTestController.setCustomPolicyDelegate() is called with only
one argument, the second ("permissive") should default to false.
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-04-15 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
fix path to layout test in new results file
https://bugs.webkit.org/show_bug.cgi?id=58618
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-15 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
fix sorting in new results file
https://bugs.webkit.org/show_bug.cgi?id=58616
-Fix the custom sort to return -1, 0, 1 instead of true/false.
-Secondary sort by test name when sort values are equal.
-Sort by test name by default.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-15 Dmitry Lomov <dslomov@google.com>
Reviewed by David Levin.
check-webkit-style shouldn't complain about not including a primary header file
if none exists
https://bugs.webkit.org/show_bug.cgi?id=39514
* Scripts/webkitpy/style/checkers/cpp.py:
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
2011-04-15 Eric Seidel <eric@webkit.org>
Reviewed by Mihai Parparita.
queues.webkit.org should display when a bot last rebooted
https://bugs.webkit.org/show_bug.cgi?id=58562
There is more repeated code here than I would like. I fear
my django-fu isn't quite up to snuff.
* QueueStatusServer/handlers/queuestatus.py:
* QueueStatusServer/templates/queuestatus.html:
2011-04-14 Alok Priyadarshi <alokp@chromium.org>
Reviewed by Adam Barth.
Remove dependency on chromium skia::PlatformCanvas
https://bugs.webkit.org/show_bug.cgi?id=57563
Fixed compile error on Mac when using Core Graphics.
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::paintRect):
2011-04-14 Alok Priyadarshi <alokp@chromium.org>
Reviewed by James Robinson.
Remove dependency on chromium skia::PlatformCanvas
https://bugs.webkit.org/show_bug.cgi?id=57563
* DumpRenderTree/chromium/TestShell.cpp:
(makeCanvasOpaque):
(TestShell::dumpImage):
* DumpRenderTree/chromium/TestShell.h:
* DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
(WebThemeControlDRTWin::WebThemeControlDRTWin):
(WebThemeControlDRTWin::draw):
(WebThemeControlDRTWin::drawTextField):
(WebThemeControlDRTWin::drawProgressBar):
* DumpRenderTree/chromium/WebThemeControlDRTWin.h:
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::canvas):
* DumpRenderTree/chromium/WebViewHost.h:
2011-04-14 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] Add -maximize flag to QtTestBrowser and MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=58007
Adds flag to start the test browsers maximized.
Symbian started maximized on QtTestBrowser - made MiniBrowser follow
suit.
Also fixes some ToggleFullScreen issues so that prior window state is
restored.
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):
(BrowserWindow::toggleFullScreenMode):
* MiniBrowser/qt/MiniBrowserApplication.cpp:
(MiniBrowserApplication::handleUserOptions):
* MiniBrowser/qt/MiniBrowserApplication.h:
(WindowOptions::WindowOptions):
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::init):
(LauncherWindow::toggleFullScreenMode):
* QtTestBrowser/launcherwindow.h:
(WindowOptions::WindowOptions):
* QtTestBrowser/main.cpp:
(LauncherApplication::handleUserOptions):
2011-04-14 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
add checkbox to only show unexpected results to new results file
https://bugs.webkit.org/show_bug.cgi?id=58606
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-13 Brent Fulgham <bfulgham@webkit.org>
Reviewed by Martin Robinson.
[WinCairo] Implement ImageDiff Logic.
http://bugs.webkit.org/show_bug.cgi?id=58486
* DumpRenderTree/win/ImageDiffCairo.cpp: Added.
(strtof):
(readFromData):
(createImageFromStdin):
(releaseMallocBuffer):
(createDifferenceImage):
(imageHasAlpha):
(writeToData):
(main):
* DumpRenderTree/win/ImageDiff.vcproj: Updated for new
ImageDiffCairo.cpp (excluding from Apple builds).
* DumpRenderTree/win/ImageDiffDebugCairoCFLite.vsprops: Use
new ImageDiffWinCairoCommon.vsprops file.
* DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops: Use
new ImageDiffWinCairoCommon.vsprops file.
* DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops: Added.
* Scripts/old-run-webkit-tests: Make sure proper executable
is built and executed for WinCairo port.
2011-04-14 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
only show expected failure type for ports that use an expectations file
https://bugs.webkit.org/show_bug.cgi?id=58588
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-14 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
commit-queue fails to catch IOError when results.html is missing
https://bugs.webkit.org/show_bug.cgi?id=58589
* Scripts/webkitpy/tool/commands/queues.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
2011-04-14 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix for Mac installation script.
* wx/packaging/build-mac-installer.py:
2011-04-14 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
sort columns in the new json_results.html file
https://bugs.webkit.org/show_bug.cgi?id=58581
Also fix padding on TDs and the path to layout tests.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html:
2011-04-14 Eric Seidel <eric@webkit.org>
Reviewed by Dimitri Glazkov.
queues.webkit.org should display when a bot last passed a patch
https://bugs.webkit.org/show_bug.cgi?id=58546
I am suspicious that some of our commit-queue instances are never
actually passing anything.
This will let us know if those bots are never passing patches.
* QueueStatusServer/handlers/queuestatus.py:
* QueueStatusServer/index.yaml:
* QueueStatusServer/templates/queuestatus.html:
2011-04-14 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
REGRESSION (r83764): webkitpy.layout_tests.port.chromium_win_unittest.ChromiumWinTest.test_default_worker_model fails on apple-windows-13
https://bugs.webkit.org/show_bug.cgi?id=58545
Remove platform-specific overrides; the default behavior should
be working everywhere now, I think.
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
2011-04-14 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
commit-queues are hitting an exception trying to land when the tree is red
https://bugs.webkit.org/show_bug.cgi?id=58558
Just an un-tested (and thus broken) code path from yesterday's changes.
* Scripts/webkitpy/common/net/layouttestresults.py:
* Scripts/webkitpy/common/net/layouttestresults_unittest.py:
* Scripts/webkitpy/tool/commands/queues.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
2011-04-14 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fixes for wxMSW and wx 2.9.1.1.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::shadowRoot):
* wx/build/wxpresets.py:
2011-04-14 Victoria Kirst <vrk@google.com>
Reviewed by James Robinson.
[chromium] Failing GPU video tests
https://bugs.webkit.org/show_bug.cgi?id=57926
Added/updated GPU directory names for rebaseline tool.
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-04-13 Daniel Bates <dbates@rim.com>
Reviewed by Adam Barth.
Perl unit test removeEOL.pl should print test case name on failure
https://bugs.webkit.org/show_bug.cgi?id=58513
* Scripts/webkitperl/VCSUtils_unittest/removeEOL.pl: Pass $title as
second argument of Test::Simple::ok().
2011-04-13 Kent Tamura <tkent@chromium.org>
Reviewed by Eric Seidel.
gdb-safari should pass arguments to safari.
https://bugs.webkit.org/show_bug.cgi?id=58439
We can do:
gdb-safari --debug LayoutTests/.../foo.html
* Scripts/gdb-safari: Passing @ARGV to Safari via gdb --arg.
2011-04-13 Dirk Pranke <dpranke@chromium.org>
Unreviewed, build fix.
https://bugs.webkit.org/show_bug.cgi?id=52763
r83799 broke test-webkitpy because it assumed we always have
access to a real filesystem (we don't during the unit tests).
Stubbing out the copy for now but we might need a better
solution at some point.
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-13 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests --run-singly option is busted
https://bugs.webkit.org/show_bug.cgi?id=55909
Python's broken lexical scoping strikes again :(.
* Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-13 Ojan Vafai <ojan@chromium.org>
Reviewed by Eric Seidel.
generated unexpected_results.html from unexpected_results.json
https://bugs.webkit.org/show_bug.cgi?id=52763
Eventually, we'll merge this with results.html and have a single richer results page.
For now, I just want to get something checked in that we can iterate on.
* Scripts/webkitpy/layout_tests/layout_package/json_results.html: Added.
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-04-13 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Up the failure limit to allow the commit-queue to actually land when the tree is red
https://bugs.webkit.org/show_bug.cgi?id=58499
This change increases the --exit-after-N-failures limit used by the commit-queue
from 1 to 10. This will cause the code added in bug 58494 to actually get
exercised and the queues should start being able to land when the trees are red.
When testing I found that test_runtests_leopard_commit_queue_hack_command was the
only unit test to actually detect this change! Since we don't actually
run on leopard anymore, I removed the hack code and repurposed the test.
* Scripts/webkitpy/tool/bot/commitqueuetask.py:
* Scripts/webkitpy/tool/steps/runtests.py:
* Scripts/webkitpy/tool/steps/steps_unittest.py:
2011-04-13 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
commit-queue should be able to land when tree is red
https://bugs.webkit.org/show_bug.cgi?id=58494
There is some yak hair on my hands, I will admit.
This change is mostly about adding an ExpectedFailures
class to track when the bots are red and we should be
ignoring failures when landing from the commit-queue.
However, to make intelligent decisions about patches we
need to know whether the run hit the --exit-after-N-failures limit
or not. Right now that information is not saved off in results.html
so we have to pull the information from RunTests.
I've plumbed the --exit-after-N-failures information into
LayoutTestResults for now to make the ExpectedFailures code cleaner.
As a result of adding all these additional calls to delegate.layout_test_results()
I broke some of our flaky test detection tests and had to re-write them
to not depend on the number of layout_test_results code.
At the same time I updated the commit-queue to use the newer filesystem
API (to allow us to use MockFileSystem) which required further changes
to the layout tests. Changes were required in either case, since
we're now calling layout_test_results() in more cases, which previously
would try and hit the disk (until I moved it to use tool.filesystem).
I should note that *all* of this code is disabled for now, since our
--exit-after-N-failures limit is currently 1! (Thus were always in the
case where we can't actually tell if the layout test results are legit.)
I will up that limit in a second patch (which may require a couple more unit test tweaks).
* Scripts/webkitpy/common/net/layouttestresults.py:
* Scripts/webkitpy/tool/bot/commitqueuetask.py:
* Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
* Scripts/webkitpy/tool/bot/expectedfailures.py: Added.
* Scripts/webkitpy/tool/bot/expectedfailures_unittest.py: Added.
* Scripts/webkitpy/tool/commands/queues.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
* Scripts/webkitpy/tool/commands/queuestest.py:
* Scripts/webkitpy/tool/steps/runtests.py:
2011-04-13 Brent Fulgham <bfulgham@webkit.org>
Unreviewed build correction.
* TestWebKitAPI/PlatformWebView.h: Limit include of
<CoreGraphics/CGGeometry> to PLATFORM(CG) users.
2011-04-13 Mihai Parparita <mihaip@chromium.org>
Reviewed by Eric Seidel.
[NRWT] Rename --baseline-search-patch to --additional-platform-directory to match ORWT
https://bugs.webkit.org/show_bug.cgi?id=58489
r83743 added --additional-platform-directory to old-run-webkit-tests,
rename the equivalent flag in new-run-webkit-tests to be consistent.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-13 Mihai Parparita <mihaip@chromium.org>
Reviewed by Eric Seidel.
Rebaseline queue server
https://bugs.webkit.org/show_bug.cgi?id=57891
Adds a simple rebaseline queue server (meant to run at
http://rebaseline-queue.appspot.com/). It presents a list of all (NRWT)
builders and their failing tests, and allows those tests to be added to
a per-builer rebaseline queue.
A subsequent change will add a mode to NRWT (and/or a wrapper script)
for pulling the tests to be rebaselined from the server and running them
with --reset-results (the initial use-case is for the GPU hardware bots,
which do not have checked-in baselines, therefore there is no need for a
check-in step yet).
* RebaselineQueueServer/app.yaml: Added.
* RebaselineQueueServer/handlers/__init__.py: Added.
* RebaselineQueueServer/handlers/builderqueue.py: Added.
* RebaselineQueueServer/handlers/pages.py: Added.
* RebaselineQueueServer/index.yaml: Added.
* RebaselineQueueServer/main.py: Added.
* RebaselineQueueServer/model/__init__.py: Added.
* RebaselineQueueServer/model/queueentry.py: Added.
* RebaselineQueueServer/static/builder-frame-empty.html: Added.
* RebaselineQueueServer/static/styles.css: Added.
* RebaselineQueueServer/templates/builder-picker.html: Added.
* RebaselineQueueServer/templates/builder-queue-edit.html: Added.
* RebaselineQueueServer/templates/builder-queue-list.html: Added.
* RebaselineQueueServer/templates/home.html: Added.
2011-04-13 Dirk Pranke <dpranke@chromium.org>
Reviewed by Mihai Parparita.
nrwt: enable multiple processes by default on Chromium Win.
https://bugs.webkit.org/show_bug.cgi?id=55163
Re-land r79268; it should be more stable now and will hopefully
work.
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2011-04-13 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: configure logging in child processes properly
https://bugs.webkit.org/show_bug.cgi?id=58296
Logging propagated properly from manager to worker processes on
Unix but not on Windows; this fixes that.
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
2011-04-13 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: suppress extraneous pretty patch warnings
https://bugs.webkit.org/show_bug.cgi?id=58295
NRWT wasn't being careful about when it logged messages from
PrettyPatch not being available; it should only log during
check_build(), and be silent otherwise. This was causing us to
get multiple errors at runtime, which was confusing.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
2011-04-13 Dirk Pranke <dpranke@chromium.org>
Reviewed by Jian Li.
add chromium-linux-x86_64 to rebaseline-chromium-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=58461
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-04-13 Tony Chang <tony@chromium.org>
Unreviewed, fix NRWT. Stack:
return super(ChromiumPort, self).results_directory(self)
TypeError: results_directory() takes exactly 1 argument (2 given)
* Scripts/webkitpy/layout_tests/port/chromium.py:
2011-04-13 Adam Roben <aroben@apple.com>
Teach ORWT about extra platform-specific directories to be searched before the default ones
Fixes <http://webkit.org/b/58446> ORWT needs a way to look in a user-specified directory for
platform-specific results
Reviewed by Anders Carlsson.
* Scripts/old-run-webkit-tests:
(top-level): Added @additionalPlatformDirectories and hooked up
--additional-platform-directories to populate it.
(buildPlatformResultHierarchy): Prepend the additional platform directories to the default
list (after filtering out any of the additional directories that don't exist).
2011-04-13 Lucas Forschler <lforschler@apple.com>
Reviewed by Adam Roben.
Add the /T command to taskkill.exe.
This will kill spawned child processes.
* BuildSlaveSupport/win/kill-old-processes:
2011-04-13 Lucas Forschler <lforschler@apple.com>
Reviewed by Adam Roben.
Sort kill list alphabeticaly.
Fixed capitalization on VcBuildHelper.
* BuildSlaveSupport/win/kill-old-processes:
2011-04-13 Lucas Forschler <lforschler@apple.com>
Reviewed by Adam Roben.
https://bugs.webkit.org/show_bug.cgi?id=58384
Update kill-old-processes to include a few new ones.
* BuildSlaveSupport/win/kill-old-processes:
2011-04-13 Gustavo Noronha Silva <gns@gnome.org>
Reviewed by Martin Robinson.
[GTK] PluginPackage should check whether a plugin mixes GTK+ 2 and 3 symbols itself
https://bugs.webkit.org/show_bug.cgi?id=58297
Do not disable flash specifically when built with GTK+ 3 -
PluginPackage should do the right thing while loading plugins.
* GtkLauncher/main.c:
(main):
2011-04-12 Kent Tamura <tkent@chromium.org>
Reviewed by Dimitri Glazkov.
[Chromium] Add WebSettings::setValidationMessageTimerMagnification()
https://bugs.webkit.org/show_bug.cgi?id=57426
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::applyTo): Disable the validation message timer.
2011-04-12 James Kozianski <koz@chromium.org>
Unreviewed.
Adding myself to committers.py.
* Scripts/webkitpy/common/config/committers.py:
2011-04-12 Brent Fulgham <bfulgham@webkit.org>
Unreviewed build correction for WinCairo bot after r83639.
https://bugs.webkit.org/show_bug.cgi?id=51790
Correct typo for the new "update-webkit-wincairo-libs" script.
* Scripts/build-webkit: Incorrectly calling the
update-webkit-cairo-libs which does not exist.
2011-04-12 Alice Liu <alice.liu@apple.com>
Reviewed by Jon Honeycutt.
A change that should have been part of http://trac.webkit.org/changeset/83628
* DumpRenderTree/win/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues): Setting should be off for most tests. Reset to false.
2011-04-12 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: --results-directory is relative to builddir, not $PWD
https://bugs.webkit.org/show_bug.cgi?id=58272
NRWT was interpreting the --results-directory cmd line arg as
relative to the build directory, not the current working
directory (ORWT uses the latter, which is much more intuitive).
This patch fixes the base case, but includes an override for
Chromium that is needed until the bots can be updated.
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-04-12 Philippe Normand <pnormand@igalia.com>
Unreviewed, roll out r83621 as it broke the GTK build.
[GTK] arguments passed to build-jsc and build-webkit scripts are not taken into account
https://bugs.webkit.org/show_bug.cgi?id=58333
* Scripts/build-jsc:
* Scripts/build-webkit:
2011-04-12 Carl Lobo <carllobo@gmail.com> and Brent Fulgham <bfulgham@webkit.org>
Reviewed by Adam Roben.
Patch to download the WinCairo dependancies as part of build-webkit.
https://bugs.webkit.org/show_bug.cgi?id=51790
This patch downloads the dependancies from
http://idisk.mac.com/bfulgham-Public/WinCairoRequirements.zip
checking the modification timestamp etc.
* Tools/Scripts/build-webkit
* Tools/Scripts/update-webkit: Modified to understand the --wincairo
flag, and to update WinCairo dependencies when used.
* Tools/Scripts/update-webkit-cairo-libs: added
* Tools/Scripts/update-webkit-dependency: added
* Tools/Scripts/update-webkit-auxiliary-libs
2011-04-12 Jeff Miller <jeffm@apple.com>
Reviewed by Adam Roben.
WebKit2: Pressing Tab in Web Inspector's console does not cycle through completion options
https://bugs.webkit.org/show_bug.cgi?id=56020
Don't call TranslateMessage() in the MiniBrowser or TestWebKitAPI for key messages destined for a WebKit2 view,
since WebKit will do this for us. If we didn't do this, TranslateMessage() would be called twice,
which would generate two characters for every keypress (for example). I didn't bother doing this for
WebKitTestRunner, since it doesn't get any WM_KEYDOWN events.
Add new WebKit2/TranslateMessageGeneratesWMChar test to test expected TranslateMessage() behavior.
* MiniBrowser/win/main.cpp:
(shouldTranslateMessage): Added.
(_tWinMain): Don't call TranslateMessage() unless shouldTranslateMessage() says to.
* TestWebKitAPI/PlatformUtilities.h: Added shouldTranslateMessage() on Windows.
* TestWebKitAPI/PlatformWebView.h: Added simulateAKeyDown().
* TestWebKitAPI/Tests/WebKit2/win/TranslateMessageGeneratesWMChar.cpp: Added.
(TestWebKitAPI::didNotHandleKeyEventCallback): Added.
(TestWebKitAPI::runAndWatchForWMChar): Added.
(TestWebKitAPI::TEST): Added.
* TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
(TestWebKitAPI::Util::run): Don't call TranslateMessage() unless shouldTranslateMessage() says to.
(TestWebKitAPI::Util::shouldTranslateMessage): Added.
* TestWebKitAPI/win/PlatformWebViewWin.cpp:
(TestWebKitAPI::PlatformWebView::simulateAKeyDown): Added.
* TestWebKitAPI/win/TestWebKitAPI.vcproj: Added TranslateMessageGeneratesWMChar.cpp.
2011-04-12 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: update code to use port.results_directory() instead of options.results_directory
https://bugs.webkit.org/show_bug.cgi?id=58290
This is some preliminary refactoring for bug 58272.
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-04-12 Dirk Pranke <dpranke@chromium.org>
Reviewed by Eric Seidel.
webkitpy: fix mock_filesystem abspath to handle relative paths
and add filesystem.chdir() and filesystem.getcwd() to be able
to test this and mock it out.
https://bugs.webkit.org/show_bug.cgi?id=58288
* Scripts/webkitpy/common/system/filesystem.py:
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/common/system/filesystem_unittest.py:
2011-04-12 Alice Liu <alice.liu@apple.com>
Reviewed by Sam Weinig.
https://bugs.webkit.org/show_bug.cgi?id=58292
Provide new setting to allow site icon loading despite disabling automatic image loading in general.
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetDefaultsToConsistentValues): Setting should be off for most tests. Reset to false.
2011-04-12 Thouraya ANDOLSI <thouraya.andolsi@st.com>
Reviewed by Eric Seidel.
[Qt] Enable JIT build for SH4 platforms.
https://bugs.webkit.org/show_bug.cgi?id=58317
* DumpRenderTree/qt/DumpRenderTree.pro:
* QtTestBrowser/QtTestBrowser.pro:
* Scripts/webkitdirs.pm:
2011-04-12 Tony Chang <tony@chromium.org>
Reviewed by Andreas Kling.
[Qt] embed checksums in PNGs written by Qt-DRT
https://bugs.webkit.org/show_bug.cgi?id=58173
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::dump):
2011-04-12 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
commit-queue should upload failure diffs when tests fail
https://bugs.webkit.org/show_bug.cgi?id=58348
This change was mostly just plumbing. We were already saving
this information for flaky test reporting. I just made it possible
for normal failures to report archives as well.
I did a little abstraction work to try and share some code between
flakytestreporter.py and this new code. There is still more we could do.
In making this change I also went through and updated the various
places we have urls hard-coded in our python and pointed them at
common.config.urls.
* Scripts/webkitpy/common/checkout/scm.py:
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/statusserver.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/tool/bot/botinfo.py: Added.
* Scripts/webkitpy/tool/bot/botinfo_unittest.py: Added.
* Scripts/webkitpy/tool/bot/commitqueuetask.py:
* Scripts/webkitpy/tool/bot/flakytestreporter.py:
* Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
* Scripts/webkitpy/tool/commands/queues.py:
2011-04-12 Daniel Bates <dbates@rim.com>
Rubber-stamped by Eric Seidel.
svn-apply and/or patch(1) has trouble applying patches that makes changes to files
with Windows line endings
https://bugs.webkit.org/show_bug.cgi?id=53625
Fixes an issue where a patch made on a SCM checkout on a Unix file system cannot be
applied to an SCM checkout on a Windows file system and vice-versa.
This issue manifests itself due to custom SCM settings as well as differences in
gitattributes support in older versions of Git (e.g. 1.7.0.3). We implement support
into svn-apply/unapply to compensate for differences in line endings by converting
the line endings in a patch to match the line endings in the target file (if it exists).
* Scripts/VCSUtils.pm:
- Added parseFirstEOL(), which returns the first end-of-line character(s) that appear in the
content read from the specified file handle. Obviously, this heuristic will not work
for files with mixed line endings, but I don't envision such files to be the norm. If
this turns out to be an issue then we can improve the heuristic.
- Added firstEOLInFile().
- Modified parsePatch() and parseDiff() to take an optional hash reference to an options hash.
In particular, added the hash key shouldNotUseIndexPathEOL to control whether to use the line
endings in the diff instead of the line endings in the target file.
- Extracted local variable $chunkRangeRegEx from fixChangeLogPatch() so that it can be
re-used in parseDiff().
* Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl: Modified to pass {shouldNotUseIndexPathEOL => 1}
to parsePatch() since these unit tests don't create mock files. Instead, created unit tests that use
mock files in file VCSUtils_unittest/parseDiffWithMockFiles.pl.
* Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl: Added.
* Scripts/webkitperl/VCSUtils_unittest/parseFirstEOL.pl: Added.
2011-04-12 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
REGRESSION (r83550): Multiple http tests crashing in the web process in SQLite beneath CFNetwork on Windows 7 Release (WebKit2 Tests)
https://bugs.webkit.org/show_bug.cgi?id=58336
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::resetStateToConsistentValues):
Remove cache clearing as that is causing tests to crash on Windows 7.
2011-04-11 Sam Weinig <sam@webkit.org>
Reviewed by Anders Carlsson.
REGRESSION(r77974): http/tests/security/aboutBlank/security-context-window-open.html is failing
<rdar://problem/8981346>
https://bugs.webkit.org/show_bug.cgi?id=54159
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage):
(WTR::InjectedBundlePage::decidePolicyForNavigationAction):
(WTR::InjectedBundlePage::decidePolicyForNewWindowAction):
(WTR::InjectedBundlePage::decidePolicyForResponse):
(WTR::InjectedBundlePage::unableToImplementPolicy):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
Add short-circuited policy client which matches WebKit1.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::resetStateToConsistentValues):
Clear the cache between tests to ensure consistent results.
2011-04-11 George Guo <George.Guo@nokia.com>
Reviewed by Eric Seidel.
prepare-ChangeLog support email input but -h did not show the option
https://bugs.webkit.org/show_bug.cgi?id=58164
Add email option to -h to improve the usability
* Scripts/prepare-ChangeLog:
2011-04-11 Dimitri Glazkov <dglazkov@chromium.org>
Reviewed by Eric Carlson.
Rename MediaControls to MediaControlRootElement.
https://bugs.webkit.org/show_bug.cgi?id=58250
* Scripts/do-webcore-rename: Documented the renaming.
2011-04-11 Sam Weinig <sam@webkit.org>
Fix Windows build.
* MiniBrowser/win/BrowserView.cpp:
(BrowserView::create):
2011-04-11 Sam Weinig <sam@webkit.org>
Reviewed by Anders Carlsson.
Move focus management out of WebKit via the UIClient
<rdar://problem/8784068>
https://bugs.webkit.org/show_bug.cgi?id=58278
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage):
(WTR::TestController::initialize):
Add stubs for new UIClient functions.
2011-02-03 Martin Robinson <mrobinson@igalia.com>
Reviewed by Eric Seidel.
[GTK] The GTK+ DRT needs an implementation of the PlainTextController
https://bugs.webkit.org/show_bug.cgi?id=53605
Add an implementation of the PlainTextController for the GTK+ DRT.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(addControllerToWindow): Added this helper method.
(webViewWindowObjectCleared): Use the new helper to add EventSender
and the PlainTextController to the window object.
* GNUmakefile.am: Added PlainTextController source files.
* DumpRenderTree/gtk/PlainTextController.cpp: Added. This will rely
on DumpRenderTreeSupportGtk to convert a JSValueRef into a WebKitDOMRange
until that functionality exists somewhere in the WebKitGTK+ stack.
* DumpRenderTree/gtk/PlainTextController.h: Added.
2011-04-11 Yael Aharon <yael.aharon@nokia.com>
Unreviewed.
Adding my IRC nick.
* Scripts/webkitpy/common/config/committers.py:
2011-04-11 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
new-run-webkit-tests: fix crash when running under windows cmd.exe
https://bugs.webkit.org/show_bug.cgi?id=58197
We would intermittently crash when running NRWT directly under
cmd.exe on windows because we were sharing the stdin file
descriptor between the python process and the http server.
cmd.exe really didn't like that, and there was no reason to
share the descriptor, so we now use a PIPE instead.
* Scripts/webkitpy/layout_tests/port/http_server.py:
2011-04-11 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: add unit tests for Port.diff_image()
https://bugs.webkit.org/show_bug.cgi?id=58196
Add the unit tests for the fix in bug 58195.
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2011-04-11 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: implement support for audio, take two
https://bugs.webkit.org/show_bug.cgi?id=58195
Attempt to re-land the fix for bug 58101 (which was initially
landed in r83330, but rolled out). This is the same patch but
fixes crashes in Port.diff_image() caused by the change for
empty image files being None instead of ''.
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/dryrun.py:
* Scripts/webkitpy/layout_tests/port/mock_drt.py:
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-04-11 Mario Sanchez Prada <msanchez@igalia.com>
Reviewed by Xan Lopez.
[GTK] Implement AccessibilityUIElement::{row|column}IndexRange in DRT
https://bugs.webkit.org/show_bug.cgi?id=57854
Implement missing features in GTK's DRT.
* DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
(indexRangeInTable): New helper function to get the range string
for the current cell inside the parent table, either from the
point of view of rows or columns.
(AccessibilityUIElement::rowIndexRange): Implemented by relying on
the new helper function indexRangeInTable().
(AccessibilityUIElement::columnIndexRange): Ditto.
2011-04-11 Mario Sanchez Prada <msanchez@igalia.com>
Reviewed by Chris Fleizach.
[GTK] Implement AccessibilityUIElement::cellForColumnAndRow in DRT
https://bugs.webkit.org/show_bug.cgi?id=57826
Implement missing feature in GTK's DRT.
* DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
(AccessibilityUIElement::cellForColumnAndRow): Implemented, by
relying on the analogous function from the AtkTable interface.
2011-04-05 Timothy Hatcher <timothy@apple.com>
Fix the extract-localizable-strings script errors that started after
WEB_UI_STRING was added to LocalizedStrings.h in WebCore.
* Scripts/extract-localizable-strings: Skip LocalizedStrings.h.
2011-04-11 Tony Chang <tony@chromium.org>
Reviewed by Eric Seidel.
fix crc computation on 64 bit machines
https://bugs.webkit.org/show_bug.cgi?id=58243
unsigned long is dependenct on CPU architecture, so use unsigned which is always 32 bits
* DumpRenderTree/CyclicRedundancyCheck.cpp:
(makeCrcTable):
(computeCrc):
* DumpRenderTree/CyclicRedundancyCheck.h:
* DumpRenderTree/PixelDumpSupport.cpp:
(appendIntToVector):
(convertChecksumToPNGComment):
2011-04-10 Kimmo Kinnunen <kimmo.t.kinnunen@nokia.com>
Reviewed by Eric Seidel.
Require no undefined symbols during compilation.
[Qt] [WK2] WebKitTestRunner, QtWebProcess and WTRInjectBundle should fail to compile when there's undefined symbols
https://bugs.webkit.org/show_bug.cgi?id=54896
Add -Wl,--no-undefined to catch missing symbols early.
* WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
* WebKitTestRunner/qt/WebKitTestRunner.pro:
2011-04-10 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix. Remove collector dirs from the list of build dirs.
* wx/build/settings.py:
2011-04-10 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix. Add Source/JavaScriptCore/heap to the list of directories.
* wx/build/settings.py:
2011-04-10 David Levin <levin@chromium.org>
Reviewed by Dimitri Glazkov.
REGRESSION(r83384): Change to chromium.py is resulting in failures of the test framework on linux.
https://bugs.webkit.org/show_bug.cgi?id=58201
* Scripts/webkitpy/layout_tests/port/chromium.py: Commented out the assert for now,
and filed bug 58202.
2011-04-09 David Levin <levin@chromium.org>
Unreviewed, rolling out r83394.
http://trac.webkit.org/changeset/83394
https://bugs.webkit.org/show_bug.cgi?id=53625
Patch was incorrect as noted in the bug.
* Scripts/svn-apply:
2011-04-09 David Levin <levin@chromium.org>
Reviewed by Adam Barth.
svn-apply and/or patch(1) has trouble applying patches that makes changes to files with Windows line endings
https://bugs.webkit.org/show_bug.cgi?id=53625
* Scripts/svn-apply: Ensure that the portions of patches
which are for vcproj/vsprops files has DOS line endings.
2011-04-09 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
It looks like NRWT has not been stopping DRT/TestShell
instances properly on windows, probably for a long time.
This would go a long way to explaining why we often have
processes lying around :)
https://bugs.webkit.org/show_bug.cgi?id=57807
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2011-04-09 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] Don't link against fontconfig or X11 if embedded
https://bugs.webkit.org/show_bug.cgi?id=58104
* DumpRenderTree/qt/DumpRenderTree.pro: fontconfig !included in embedded
* QtTestBrowser/QtTestBrowser.pro: ditto
* WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: ditto
* WebKitTestRunner/qt/WebKitTestRunner.pro: ditto
2011-04-08 Jon Honeycutt <jhoneycutt@apple.com>
http/tests/loading/preload-append-scan.php is failing on Windows
https://bugs.webkit.org/show_bug.cgi?id=58178
Reviewed by Brian Weinstein.
* DumpRenderTree/win/DumpRenderTree.cpp:
(lastPathComponent):
Use CFURL API to find the last path component, rather than
PathFindFileName(), because the latter will include the query string.
2011-04-08 Jian Li <jianli@chromium.org>
Unreviewed, rolling out r83327, r83330 since these patches are very
likely to break chromium webkit mac10.6 builders.
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/dryrun.py:
* Scripts/webkitpy/layout_tests/port/mock_drt.py:
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-04-08 Vincent Scheib <scheib@chromium.org>
Reviewed by Eric Seidel.
add chromium-gpu-linux-x86_64 to the recognized list of ports
https://bugs.webkit.org/show_bug.cgi?id=58099
Add chromium-gpu-linux-x86_64 to fix assert.
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2011-04-08 Alpha Lam <hclam@chromium.org>
Unreviewed, rolling out r83335.
http://trac.webkit.org/changeset/83335
https://bugs.webkit.org/show_bug.cgi?id=53556
GTK and QT bots are broken
* Scripts/build-webkit:
2011-04-07 Anna Cavender <annacc@chromium.org>
Reviewed by Eric Carlson.
Setup ENABLE(TRACK) feature define
https://bugs.webkit.org/show_bug.cgi?id=53556
* Scripts/build-webkit:
2011-04-08 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
new-run-webkit-tests: configure a NRWT test bot for Mac SL
https://bugs.webkit.org/show_bug.cgi?id=58114
This change removes the old new-run-webkit-tests entry and
updates it for a new test-only bot.
* BuildSlaveSupport/build.webkit.org-config/config.json:
2011-04-08 Dominic Cooney <dominicc@google.com>
Reviewed by Adam Roben.
Make layoutTestController.shadowRoot return null, not undefined,
when its argument is invalid.
https://bugs.webkit.org/show_bug.cgi?id=58121
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::shadowRoot):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::shadowRoot):
2011-04-08 Pere Martir <pere.martir4@gmail.com>
Reviewed by Adam Roben.
Locate NSTD.EXE in 64-bit Windows
https://bugs.webkit.org/show_bug.cgi?id=57847
* Scripts/old-run-webkit-tests:
2011-04-08 Adam Roben <aroben@apple.com>
Qt build fix
* DumpRenderTree/TestNetscapePlugIn/PluginObject.h: Pull in stdarg.h for va_list.
2011-04-07 Adam Roben <aroben@apple.com>
Test that NPP_SetWindow is passed a null window handle during plugin destruction on non-Mac platforms
Test for <http://webkit.org/b/47009> WebKit2 needs to call NPP_SetWindow when destroying a
plugin
Reviewed by Anders Carlsson.
* DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
(pluginLogWithArguments): Moved code to format and log the message here...
(pluginLog): ...from here.
* DumpRenderTree/TestNetscapePlugIn/PluginObject.h: Added pluginLogWithArguments.
* DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
(PluginTest::log): Added. Calls through to pluginLogWithArguments.
* DumpRenderTree/TestNetscapePlugIn/PluginTest.h: Added log.
* DumpRenderTree/TestNetscapePlugIn/Tests/NPPSetWindowCalledDuringDestruction.cpp: Added.
(NPPSetWindowCalledDuringDestruction::setWillBeDestroyed): Records that destruction is about
to begin.
(NPPSetWindowCalledDuringDestruction::NPPSetWindowCalledDuringDestruction): Simple
constructor.
(NPPSetWindowCalledDuringDestruction::NPP_GetValue): Creates and returns a ScriptObject that
can be used to invoke our setWillBeDestroyed function.
(NPPSetWindowCalledDuringDestruction::NPP_SetWindow): Records what has happened (and logs if
anything unexpected happens).
(NPPSetWindowCalledDuringDestruction::NPP_Destroy): On Mac, logs a failure message if
NPP_SetWindow was called during destruction. On other platforms, logs a failure message if
NPP_SetWindow was *not* called during destruction.
(NPPSetWindowCalledDuringDestruction::ScriptObject::hasMethod): Return true for our only
method, setWillBeDestroyed.
(NPPSetWindowCalledDuringDestruction::ScriptObject::invoke): Call through to the PluginTest
object.
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* GNUmakefile.am:
Added new test.
* Scripts/old-run-webkit-tests: Skip the new test when using out-of-process plugins with
WebKit1 on Mac, since it can't work properly due to <http://webkit.org/b/58077>.
2011-04-08 Mario Sanchez Prada <msanchez@igalia.com>
Reviewed by Martin Robinson.
[GTK] Implement increment() and decrement() functions in DRT's AccessibilityUIElement
https://bugs.webkit.org/show_bug.cgi?id=58039
Implement missing functions in GTK's DRT.
* DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
(AccessibilityUIElement::increment): Implemented.
(AccessibilityUIElement::decrement): Implemented.
2011-04-08 Dominic Cooney <dominicc@google.com>
Reviewed by Kent Tamura.
Make WK2 layoutTestController.shadowRoot return undefined, not
null, when its argument is not an element.
https://bugs.webkit.org/show_bug.cgi?id=58121
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::shadowRoot):
2011-04-07 Dirk Pranke <dpranke@chromium.org>
Reviewed by Mihai Parparita.
Change reference port for Mac GPU baselines from
Leopard to SnowLeopard, XP to Win7, and Linux-x86
to Linux x86-64.
https://bugs.webkit.org/show_bug.cgi?id=58099
* Scripts/webkitpy/layout_tests/port/chromium.py:
2011-04-07 Andrew Scherkus <scherkus@chromium.org>
Revert ENABLE_TRACK patch due to compile failures.
* Scripts/build-webkit:
2011-04-07 Tony Chang <tony@chromium.org>
Reviewed by Eric Seidel.
update DRT to embed checksums in png files
https://bugs.webkit.org/show_bug.cgi?id=57871
We insert the bytes for the comment in printPNG rather than at encode
time because each platform does its own PNG encoding (either using CG
or cairo). Putting this in pringPNG avoids having to duplicate this
code, although it's not as clean as doing it at encoding time.
We insert the comment right after the IHDR chunk of the PNG.
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/CyclicRedundancyCheck.cpp: Added.
(makeCrcTable): Ported from LayoutTests/fast/canvas/webgl/resources/pnglib.js
(computeCrc): Ported from LayoutTests/fast/canvas/webgl/resources/pnglib.js
* DumpRenderTree/CyclicRedundancyCheck.h: Added.
* DumpRenderTree/PixelDumpSupport.cpp:
(dumpWebViewAsPixelsAndCompareWithExpected):
(convertChecksumToPNGComment): Generate the bytes to insert.
(printPNG): Insert the png comment before the first IDAT section.
* DumpRenderTree/PixelDumpSupport.h:
* DumpRenderTree/cairo/PixelDumpSupportCairo.cpp:
(printPNG):
(dumpBitmap):
* DumpRenderTree/cg/PixelDumpSupportCG.cpp:
(printPNG):
(dumpBitmap):
* DumpRenderTree/win/DumpRenderTree.vcproj:
* GNUmakefile.am:
2011-04-07 Jeff Miller <jeffm@apple.com>
Reviewed by Adam Roben.
Replace WKStringGetCharactersPtr() with WKStringGetCharacters()
https://bugs.webkit.org/show_bug.cgi?id=58058
* TestWebKitAPI/Tests/WebKit2/WKString.cpp:
(TestWebKitAPI::TEST): Add tests for WKStringGetLength() and WKStringGetCharactersPtr().
2011-04-07 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.isPageBoxVisible
https://bugs.webkit.org/show_bug.cgi?id=42695
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::isPageBoxVisible):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-07 Adam Roben <aroben@apple.com>
Build fix
* BuildSlaveSupport/build.webkit.org-config/master.cfg: Added a missing import.
2011-04-07 Adam Roben <aroben@apple.com>
Move ExtractTestResults[AndLeaks]'s summarizing code back from commandComplete() to finished()
commandComplete() is never called for MasterShellCommands like ExtractTestResults[AndLeaks].
(Unfortunately the buildbot documentation does not make this clear.) finished() is the only
hook we have, so we have to do our work there. I added a new addCustomURLs method which can
be overridden by subclasses to provide extra URLs before we call up to the base class (after
which adding more URLs is no longer possible).
Really hopefully fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks
bot results page
Reviewed by John Sullivan.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(ExtractTestResults.addCustomURLs): Moved code to add the "view results" URL here...
(ExtractTestResults.finished): ...from here. This replaces commandComplete, since that
method is never called for MasterShellCommands.
(ExtractTestResultsAndLeaks.addCustomURLs): Replaced commandComplete (which is never called)
with this method (which is).
2011-04-07 Adam Roben <aroben@apple.com>
Move ExtractTestResults[AndLeaks]'s summarizing code from finished() to commandComplete()
This matches how most of our other build steps work, so is good just for improving
consistency between build steps. It should also make it possible for
ExtractTestResultsAndLeaks to successfully add a URL to Leaks Viewer. (Previously we were
trying to do this in finished() after we had called up to the base class, but that was
apparently too late to add more URLs.)
Hopefully fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks bot
results page
Reviewed by John Sullivan.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(ExtractTestResults): Added a descriptionDone property so that we don't have to manually
call setText to get the right text to show up for this step. This matches how most of our
other build steps work.
(ExtractTestResults.commandComplete): Replaced our override of finished with this function.
This is how most of our other build steps work.
(ExtractTestResultsAndLeaks.commandComplete): Replaced our override of finished with this
function, to match the base class.
2011-04-07 Adam Barth <abarth@webkit.org>
Reviewed by Maciej Stachowiak.
webkit-patch shouldn't state the whole working copy when calling check-webkit-style
https://bugs.webkit.org/show_bug.cgi?id=58022
Previously, we were ignoring the args variable! This patch fixes the
regression introduced in http://trac.webkit.org/changeset/82771.
* Scripts/webkitpy/tool/commands/download_unittest.py:
* Scripts/webkitpy/tool/mocktool.py:
* Scripts/webkitpy/tool/steps/checkstyle.py:
2011-04-07 Maciej Stachowiak <mjs@apple.com>
Reviewed by Adam Barth.
REGRESSION: "webkit-patch land" doesn't work correctly in SVN subdirectories
https://bugs.webkit.org/show_bug.cgi?id=58017
* Scripts/webkitpy/common/checkout/scm.py:
* Scripts/webkitpy/common/checkout/scm_unittest.py:
2011-04-06 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.pageSizeAndMarginsInPixels
https://bugs.webkit.org/show_bug.cgi?id=57984
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::pageSizeAndMarginsInPixels):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-06 Chris Rogers <crogers@google.com>
Reviewed by Tony Chang.
Add web audio support to DumpRenderTree (mac port)
https://bugs.webkit.org/show_bug.cgi?id=57969
* DumpRenderTree/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(setEncodedAudioDataCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
(LayoutTestController::dumpAsAudio):
(LayoutTestController::setDumpAsAudio):
(LayoutTestController::encodedAudioData):
(LayoutTestController::setEncodedAudioData):
* DumpRenderTree/mac/DumpRenderTree.mm:
(dumpAudio):
(dump):
2011-04-06 Benjamin Poulain <benjamin.poulain@nokia.com>
Reviewed by Kenneth Rohde Christiansen.
[Qt] LayoutTestController needs to implement numberOfPendingGeolocationPermissionRequests
https://bugs.webkit.org/show_bug.cgi?id=56086
Add the missing method.
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::getAllPages): Add a method to return all the pages allocated for the current test.
* DumpRenderTree/qt/DumpRenderTreeQt.h:
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
2011-04-06 Dai Mikurube <dmikurube@chromium.org>
Reviewed by David Levin.
Add QUOTA build flag for unified quota API
https://bugs.webkit.org/show_bug.cgi?id=57918
* Scripts/build-webkit: Added QUOTA build flag
2011-04-06 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
Adjust the apple webkit port's default timeout to match
old-run-webkit-tests at 35 seconds.
https://bugs.webkit.org/show_bug.cgi?id=37738
* Scripts/webkitpy/layout_tests/port/mac.py:
2011-04-06 Scott Cameron <sccameron@rim.com>
Reviewed by Eric Seidel.
Remove global variable $httpdPath and replace with the return value of
getHTTPDPath(). Also remove unnecessary calls to getHTTPDPath().
https://bugs.webkit.org/show_bug.cgi?id=53499
* Scripts/run-iexploder-tests:
* Scripts/webkitperl/httpd.pm:
2011-04-06 Kevin Ollivier <kevino@theolliviers.com>
Reviewed by Darin Adler.
Make sure JS_EXPORT_PRIVATE is an empty define when we aren't using the export macros.
https://bugs.webkit.org/show_bug.cgi?id=27551
* DumpRenderTree/config.h:
* WebKitAPITest/config.h:
* WebKitTestRunner/config.h:
2011-04-06 Tony Chang <tony@chromium.org>
Reviewed by Darin Adler.
teach run-webkit-tests to read checksums from png files
https://bugs.webkit.org/show_bug.cgi?id=57993
* Scripts/old-run-webkit-tests: Read the first 2k of a .png if there's
no .checksum and look for the checksum in there.
2011-04-06 David Dorwin <ddorwin@chromium.org>
Reviewed by David Levin.
Enable fullscreen layout tests for Chromium
https://bugs.webkit.org/show_bug.cgi?id=55726
Make DumpRenderTree always run with fullscreen enabled (equivalent of --enable-fullscreen).
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::applyTo):
2011-04-06 Adam Roben <aroben@apple.com>
Add a "view leaks" link to builds on SnowLeopard Intel Leaks
Fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks bot results page
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(ExtractTestResults.resultDirectoryURL): Added. Moved code to calculate the URL for the
build's results directory here...
(ExtractTestResults.finished): ...from here.
(ExtractTestResultsAndLeaks): New class that's used by the leaks builder
(ExtractTestResultsAndLeaks.finished): Calls up to the base class, but also adds a "view
leaks" link to point to Leaks Viewer for this build.
(TestFactory): Added ExtractTestResultsClass abstraction. This isn't overridden anywhere,
but it seemed good to add for consistency with BuildAndTestFactory.
(BuildAndTestFactory): Added ExtractTestResultsClass abstraction.
(BuildAndTestLeaksFactory): Use ExtractTestResultsAndLeaks as our ExtractTestResultsClass so
that we'll get a "view leaks" link.
2011-04-06 Zan Dobersek <zandobersek@gmail.com>
Reviewed by Eric Seidel.
[Gtk] plugins/set-status.html fails
https://bugs.webkit.org/show_bug.cgi?id=57844
Allow an empty status text to be dumped.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(webViewStatusBarTextChanged):
2011-04-06 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.pageNumberForElementById
https://bugs.webkit.org/show_bug.cgi?id=42329
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::pageNumberForElementById):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-06 Sergio Villar Senin <svillar@igalia.com>
Reviewed by Martin Robinson.
[GTK] DumpRenderTree: do not try to free NULL SoupURIs
https://bugs.webkit.org/show_bug.cgi?id=57932
Some Layout tests have invalid URIs that do not generate valid
SoupURI instances. Do not try to free those NULL SoupURIs.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(willSendRequestCallback):
2011-04-06 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Andreas Kling.
ORWT shouldn't generate diff files for tests without expected files
https://bugs.webkit.org/show_bug.cgi?id=57846
* Scripts/old-run-webkit-tests:
2011-04-05 Kevin Ollivier <kevino@theolliviers.com>
[wx] Mac build fix. Handle cases where 10.4 SDK is not installed, and also add x86_64 arch
to deps.
* wx/install-unix-extras:
2011-04-05 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r82978, r82999, and r83001.
http://trac.webkit.org/changeset/82978
http://trac.webkit.org/changeset/82999
http://trac.webkit.org/changeset/83001
https://bugs.webkit.org/show_bug.cgi?id=57913
Does not work in Python 2.5 (Requested by abarth on #webkit).
* Scripts/webkitpy/common/system/executive.py:
* Scripts/webkitpy/common/system/executive_unittest.py:
2011-04-05 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.numberOfPages
https://bugs.webkit.org/show_bug.cgi?id=42694
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::numberOfPages):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-05 Tony Chang <tony@chromium.org>
Reviewed by Mihai Parparita.
[chromium] stop putting results downloaded from WebKit Linux in chromium-linux-x86_64
https://bugs.webkit.org/show_bug.cgi?id=57889
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-04-05 Adam Barth <abarth@webkit.org>
Silly with statement, from the future!
* Scripts/webkitpy/common/system/executive.py:
2011-04-05 Adam Barth <abarth@webkit.org>
Reviewed by Tony Chang.
Don't use Exception.message because it's deprecated
https://bugs.webkit.org/show_bug.cgi?id=57892
Suppress the warning for now. When we move to Python 3, we might need
to something more dramatic.
* Scripts/webkitpy/common/system/executive.py:
2011-04-05 Mihai Parparita <mihaip@chromium.org>
Reviewed by Ojan Vafai.
Add builders.js to dashboard file list
https://bugs.webkit.org/show_bug.cgi?id=57899
Add file added by http://crrev.com/80538 to dashboard file list.
* TestResultServer/handlers/dashboardhandler.py:
2011-04-05 MORITA Hajime <morrita@google.com>
Reviewed by Adam Barth.
webkit-patch should print git's stderr when git svn dcommit fail
http://webkit.org/b/57861
* Scripts/webkitpy/common/checkout/scm.py:
2011-04-05 Mihai Parparita <mihaip@chromium.org>
Reviewed by Tony Chang.
Add --baseline-search-path to NRWT
https://bugs.webkit.org/show_bug.cgi?id=56233
Add NRWT option to specify additional directories to look for baselines
(will be used by hardware GPU bots which will have local per-bot
expectations for some tests)
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-04-05 Adam Roben <aroben@apple.com>
Strip off /results.html from results URLs before trying to load leaks files from them
r82734 changed build.webkit.org's "view results" URLs to point straight to the results.html
files, rather than pointing to the directory that contains them. This is more convenient for
people browsing build.webkit.org, but confused Leaks Viewer.
Fixes <http://webkit.org/b/57869> REGRESSION (r82734): Links in Leaks Viewer's Recent Builds
list don't work
Reviewed by Joseph Pecoraro.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
(RecentBuildsLoader.prototype.start): Strip off "/results.html" from the results URL.
2011-04-05 Carol Szabo <carol@webkit.org>
Unreviewed.
Updated my info in committers.py
Scripts\webkitpy\common\config\committers.py
2011-04-05 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Laszlo Gombos.
[Qt] Make WebKitLibraries optional for building QtWebKit
https://bugs.webkit.org/show_bug.cgi?id=57542
* Scripts/build-webkit:
2011-04-05 Chang Shu <cshu@webkit.org>
Reviewed by Laszlo Gombos.
[Qt] MiniBrowser defaultUrl does not work
https://bugs.webkit.org/show_bug.cgi?id=57021
Match the behavior and coding of MiniBrowser to QtTestBrowser.
* MiniBrowser/qt/main.cpp:
(main):
2011-04-05 Jade Han <jade.han@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] [Symbian] Disable WebKitTestRunner for Symbian
https://bugs.webkit.org/show_bug.cgi?id=54977
This change is a preparation to enable building webkit2 for Symbian.
Bug 57834 is filed to fix and enable WebKitTestRunner for Symbian.
* Tools.pro:
2011-04-05 Zoltan Horvath <zoltan@webkit.org>
Reviewed by Andreas Kling.
[Qt] Fix timeoutTimer of MiniBrowser's UrlLoader
https://bugs.webkit.org/show_bug.cgi?id=57832
Only QWKPage has loadFinished signal so connect to it instead of BrowserWindow.
* MiniBrowser/qt/UrlLoader.cpp:
(UrlLoader::UrlLoader):
2011-04-05 Zoltan Horvath <zoltan@webkit.org>
[Qt] Linux Release minimal build fix after r82919.
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::loadURLListFromFile):
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::loadURLListFromFile):
2011-04-05 Zoltan Horvath <zoltan@webkit.org>
Reviewed by Andreas Kling.
[Qt] Make URL loader accessible from testbrowser's menus
https://bugs.webkit.org/show_bug.cgi?id=57823
Add "Load URLs from file" to QtTestBrowser's and to MiniBrowser's menu.
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):
(BrowserWindow::loadURLListFromFile):
(BrowserWindow::~BrowserWindow):
* MiniBrowser/qt/BrowserWindow.h:
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::LauncherWindow):
(LauncherWindow::~LauncherWindow):
(LauncherWindow::createChrome):
(LauncherWindow::loadURLListFromFile):
* QtTestBrowser/launcherwindow.h:
2011-04-04 Sam Weinig <sam@webkit.org>
Reviewed by Brian Weinstein.
Remove duplicate API from WKContext
<rdar://problem/8727879>
https://bugs.webkit.org/show_bug.cgi?id=57815
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKit2/SendingMessagesToTheWebProcessBeforeItIsValid.cpp: Removed.
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
Remove SendingMessagesToTheWebProcessBeforeItIsValid since the API it was testing is now removed.
2011-04-04 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[chromium] don't write .checksum files if a fallback platform has an embedded checksum
https://bugs.webkit.org/show_bug.cgi?id=57783
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
2011-04-04 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix, run uninstall when cleaning to remove built files
from WebKitBuild.
* Scripts/webkitdirs.pm:
2011-04-04 Martin Robinson <mrobinson@igalia.com>
Reviewed by Gustavo Noronha Silva.
[GTK] WebGL support
https://bugs.webkit.org/show_bug.cgi?id=31517
Add support to the DRT for turning on WebGL when a layout tests requests it.
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::overridePreference): Allow turning on WebGL from tests.
2011-04-04 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.setAllowFileAccessFromFileURLs
https://bugs.webkit.org/show_bug.cgi?id=57572
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setAllowFileAccessFromFileURLs):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-04 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Csaba Osztrogonác.
[Qt] DumpRenderTree breaks compilation in some uClibc environments
https://bugs.webkit.org/show_bug.cgi?id=57602
* DumpRenderTree/qt/main.cpp:
(get_backtrace):
2011-04-04 Kevin Ollivier <kevino@theolliviers.com>
[wx] Unreviewed build fix, add new LayoutTestController method stub to wx.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::shadowRoot):
2011-04-04 Pavel Podivilov <podivilov@chromium.org>
Unreviewed, fix exception in rebaseline tool.
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-04-02 Beth Dakin <bdakin@apple.com>
Rubber-stamped by Geoff Garen.
For Dan!
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
2011-04-02 Beth Dakin <bdakin@apple.com>
Rubber-stamped by Geoff Garen.
Need to reset the scale, much like zoom.
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
2011-04-02 Sam Weinig <sam@webkit.org>
Reviewed by Beth Dakin.
https://bugs.webkit.org/show_bug.cgi?id=57605
Frame::pageScaleFactor() should not affect getBoundingClientRect() or
getClientRects()
-and corresponding-
<rdar://problem/9194541>
Add DRT support for the scaleWebView SPI.
* DumpRenderTree/mac/EventSendingController.mm:
(+[EventSendingController isSelectorExcludedFromWebScript:]):
(+[EventSendingController webScriptNameForSelector:]):
(-[EventSendingController scalePageBy:atX:andY:]):
* WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
* WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
(WTR::EventSendingController::scalePageBy):
* WebKitTestRunner/InjectedBundle/EventSendingController.h:
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::reset):
2011-04-02 Dominic Cooney <dominicc@google.com>
Reviewed by Martin Robinson.
Add layoutTestController.shadowRoot to GTK DumpRenderTree.
https://bugs.webkit.org/show_bug.cgi?id=57551
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::shadowRoot):
2011-04-02 Patrick Gansterer <paroga@webkit.org>
Reviewed by Eric Seidel.
Remove AbstractStep._run_script and move script names to ports.py
https://bugs.webkit.org/show_bug.cgi?id=57704
Replace deprecated _run_script with _tool.executive.run_and_throw_if_fail.
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/tool/commands/download_unittest.py:
* Scripts/webkitpy/tool/commands/roll_unittest.py:
* Scripts/webkitpy/tool/commands/upload_unittest.py:
* Scripts/webkitpy/tool/steps/abstractstep.py:
* Scripts/webkitpy/tool/steps/checkstyle.py:
* Scripts/webkitpy/tool/steps/preparechangelog.py:
* Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
* Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
2011-04-02 Patrick Gansterer <paroga@webkit.org>
Reviewed by Eric Seidel.
Emulate shebang on Win32
https://bugs.webkit.org/show_bug.cgi?id=55927
Scripts on Windows work only if they are called with the explicit interpreter.
Read the first line of scripts to detect the correct executable.
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/common/system/executive.py: Added interpreter_for_script().
* Scripts/webkitpy/common/system/executive_unittest.py:
2011-04-01 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
apos entities shouldn't show up in ChangeLogs when using webkit-patch
https://bugs.webkit.org/show_bug.cgi?id=57692
Previously, we were using BeautifulSoup to process XML from
bugs.webkit.org, but that's incorrect. We should be using
BeautifulStoneSoup to process the XML. We were getting the &apos;
entity wrong because &apos; is an XML entity but not an HTML entity.
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
2011-04-01 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r82721.
http://trac.webkit.org/changeset/82721
https://bugs.webkit.org/show_bug.cgi?id=57687
This patch introduced assertion failures on the GTK+ bots.
(Requested by mrobinson on #webkit).
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::shadowRoot):
2011-04-01 Csaba Osztrogonác <ossy@webkit.org>
Unreviewed buildfix.
[Qt][WK2] Build Webkit2 using "-2" option on Qt
https://bugs.webkit.org/show_bug.cgi?id=55074
* Scripts/build-webkit: Ensure that "-2" isn't passed to qmake.
(The isWK2() function removes it from @ARGV, but not from @options.)
2011-04-01 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Csaba Osztrogonác.
[Qt] [WK2] MiniBrowser.qrc not found - regression from rev 82671
https://bugs.webkit.org/show_bug.cgi?id=57666
* MiniBrowser/qt/MiniBrowser.qrc: Renamed from Tools/MiniBrowser/MiniBrowser.qrc.
2011-04-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
new-run-webkit-tests: fix feature detection, skipped platform lists on mac
We apparently never implemented the code to skip tests based on
what was compiled into DRT. Also, change the logic used to skip
platform directories to match what old-run-webkit-tests does:
skip every test not in a directory in the baseline search path.
https://bugs.webkit.org/show_bug.cgi?id=57662
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
2011-03-29 Mark Rowe <mrowe@apple.com>
Reviewed by Jon Honeycutt.
<http://webkit.org/b/56730> new-run-webkit-tests fails on Lion seed
Teach new-run-webkit-tests about the concept of an unreleased version of Mac OS X.
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2011-04-01 Chang Shu <cshu@webkit.org>
Reviewed by Csaba Osztrogonác.
[Qt][WK2] Build Webkit2 using "-2" option on Qt
https://bugs.webkit.org/show_bug.cgi?id=55074
* Scripts/build-webkit:
2011-04-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: remove spurious port version override in webkit.py
base.py provides a default implementation so this is just
breaking things.
https://bugs.webkit.org/show_bug.cgi?id=57667
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2011-04-01 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Adam Roben.
Make view results on waterfall direct link to results.html
https://bugs.webkit.org/show_bug.cgi?id=57671
* BuildSlaveSupport/build.webkit.org-config/master.cfg: Add "/results.html" to the URL.
* Scripts/old-run-webkit-tests: Add links to httpd access and error logs.
2011-04-01 Sam Weinig <sam@webkit.org>
Fix windows build.
* TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp:
(TestWebKitAPI::TEST):
* TestWebKitAPI/Tests/WebKit2/win/ResizeViewWhileHidden.cpp:
(TestWebKitAPI::flushMessages):
2011-04-01 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
Add adoptWK to WKRetainPtr.h
https://bugs.webkit.org/show_bug.cgi?id=57670
* TestWebKitAPI/PlatformUtilities.h:
* TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp:
* TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
* TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp:
* TestWebKitAPI/Tests/WebKit2/CookieManager.cpp:
* TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp:
* TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
* TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp:
* TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
(TestWebKitAPI::createSessionStateContainingFormData):
* WebKitTestRunner/StringFunctions.h:
Replace custom versions of adoptWK with the API on in WebKit2/WKRetainPtr.h.
2011-04-01 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
Fix leak noticed by Adam Roben in LayoutTestController::shadowRoot.
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::shadowRoot):
Make judicious use of adoptWK().
2011-04-01 Sam Weinig <sam@webkit.org>
Reviewed by Timothy Hatcher.
Fix extract-localizable-strings for macro change from UI_STRING -> WEB_UI_STRING.
* Scripts/extract-localizable-strings:
2011-04-01 Anders Carlsson <andersca@apple.com>
Reviewed by Sam Weinig.
Shockwave plug-in doesn't accept mouse events
https://bugs.webkit.org/show_bug.cgi?id=57653
<rdar://problem/8483273>
Add a plug-in test.
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
(PluginTest::indicateTestFailure):
Move code from NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject here.
(PluginTest::NPN_ConvertPoint):
* DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
* DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp:
(NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
Call indicateTestFailure.
* DumpRenderTree/TestNetscapePlugIn/Tests/mac: Added.
* DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp: Added.
(ConvertPoint::ConvertPoint):
(ConvertPoint::testConvert):
(ConvertPoint::NPP_New):
2011-04-01 Dominic Cooney <dominicc@google.com>
Reviewed by Martin Robinson.
Add layoutTestController.shadowRoot to GTK DumpRenderTree.
https://bugs.webkit.org/show_bug.cgi?id=57551
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::shadowRoot):
2011-04-01 Sam Weinig <sam@webkit.org>
Reviewed by Anders Carlsson.
WebKitTestRunner needs layoutTestController.shadowRoot
https://bugs.webkit.org/show_bug.cgi?id=57661
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::shadowRoot):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
Implement layoutTestController.shadowRoot for WebKit2.
2011-04-01 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.setDatabaseQuota
https://bugs.webkit.org/show_bug.cgi?id=57568
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setDatabaseQuota):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-04-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
This patch completes the cleanup of
rebaseline-chromium-webkit-tests to work with all of the
variants of a platform (we can now rebaseline gpu- and non-gpu
files at the same time).
When the rebaselining is complete, any lines declared as
REBASELINE in the expectations file that matches a test that was
actually rebaselined will be deleted, even if only one of the
variants was actually rebaselined. This may cause odd problems,
but is better than where we're at today.
This change removes the -g flag and deprecates -w. The -g flag is gone
because GPU baselines are handled just like any other variant.
The -w flag is deprecated because this tool now only works
against the canaries, since that's the only place we have a full
set of bots. It will be trivial to change this to
build.webkit.org if we decide that's where we want them to be.
Also, this patch deletes a lot of cruft that is no longer needed
in the test_expectations code and the port-specific code.
https://bugs.webkit.org/show_bug.cgi?id=55191
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/webkit.py:
* Scripts/webkitpy/layout_tests/port/win.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
2011-04-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
rebaseline-chromium-webkit-tests does not work correctly with
version-specific baselines. This patch updates the tool to use
all of the version-specific bots on the canaries, and will now
attempt to rebaseline all of the versions by default, although
it will not update both GPU and CPU versions.
Also, it will no longer modify the test_expectations.txt file
*at all*. You will have to manually delete the REBASELINE lines
after running the tool and determining that it did what you
wanted it to do. This should be fixed in a separate bug - see
webkit bug #55191.
https://bugs.webkit.org/show_bug.cgi?id=55608
* Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
* Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
2011-03-31 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[chromium] update the rebaseline tool to know about pngs with checksums
https://bugs.webkit.org/show_bug.cgi?id=57481
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Skip over .checksum files if the checksum is already in the png
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
2011-04-01 Adam Roben <aroben@apple.com>
Retrieve revision numbers from the build's got_revision property in Leaks Viewer
Previously, we were getting the revision of the first revision that triggered a build.
Choosing the last revision would have been more accurate. But got_revision is what is used
everywhere else on build.webkit.org, and should work even when there were no changes that
triggered a build (e.g., if someone clicked the Force Build button).
Fixes <http://webkit.org/b/57630> Leaks viewer gets some revision numbers wrong in the
recent builds list
Reviewed by Anders Carlsson.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
(RecentBuildsLoader.prototype.start): Pull the revision number out of the got_revision
property, rather than out of the first (i.e., earliest) change in the sourceStamp object.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
(Array.prototype.first): Added this helper function to return the first element in an array
that matches the given predicate, or null if no such element exists.
2011-03-31 Adam Roben <aroben@apple.com>
Prefer (but don't require) bug URLs to be on their own line when parsing bug numbers from ChangeLogs
Fixes <http://webkit.org/b/57579> webkit-patch is too strict about bug URL formatting
Reviewed by Darin Adler.
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
(parse_bug_id_from_changelog): Fall back to parse_bug_id if we weren't able to find a bug
URL on its own line.
* Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
(BugzillaTest.test_parse_bug_id_from_changelog): Updated expected results for test
progression, and added a new test that uses a short bug URL while I was at it.
2011-04-01 Adam Roben <aroben@apple.com>
Mark .vcproj/.vsprops/.sln files as being Windows-only
Fixes <http://webkit.org/b/57489> Mac builders built 82512, but shouldn't have
Reviewed by Anders Carlsson.
* Scripts/webkitpy/common/config/build.py:
(_should_file_trigger_build): Added patterns to mark .vcproj/.vsprops/.sln files and .vcproj
directories as Windows-only.
* Scripts/webkitpy/common/config/build_unittest.py:
(ShoulBuildTest): Added test cases for the above.
2011-04-01 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] Build MiniBrowser for Symbian
https://bugs.webkit.org/show_bug.cgi?id=56319
Have MiniBrowser reference it's own copy of useragentlist.txt instead
of copying QtTestBrowser's.
Remove Tools/MiniBrowser/DerivedSources.pro
This reduces complexity in the Tools scripts due to Symbian limitations.
* DerivedSources.pro:
* MiniBrowser/DerivedSources.pro: Removed.
* MiniBrowser/MiniBrowser.qrc:
* MiniBrowser/qt/MiniBrowser.pro:
* Scripts/webkitdirs.pm:
2011-03-31 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.clearAllDatabases
https://bugs.webkit.org/show_bug.cgi?id=42540
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::clearAllDatabases):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-03-31 Chang Shu <cshu@webkit.org>
Reviewed by Darin Adler.
WebKitTestRunner needs layoutTestController.setAllowUniversalAccessFromFileURLs
https://bugs.webkit.org/show_bug.cgi?id=42692
* WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2011-03-31 Darin Adler <darin@apple.com>
Reviewed by Sam Weinig.
Implement mouseDown, mouseUp, and mouseMoveTo in WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=57573
* WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
Added real definitions for mouseDown, mouseUp, mouseMoveTo and leapForward.
Removed fake definitions of keyDown and contextClick.
* WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
(WTR::operator==): Added. So we can compare two WKPoint structs.
(WTR::parseModifier): Added.
(WTR::parseModifierArray): Added.
(WTR::EventSendingController::EventSendingController): Initialize the
new data members.
(WTR::EventSendingController::mouseDown): Added. Calls
WKBundlePageSimulateMouseDown.
(WTR::EventSendingController::mouseUp): Added. Calls
WKBundlePageSimulateMouseUp.
(WTR::EventSendingController::mouseMoveTo): Added. Calls
WKBundlePageSimulateMouseMotion.
(WTR::EventSendingController::leapForward): Added.
(WTR::EventSendingController::updateClickCount): Added. Used by the
mouseDown/Up functions to create a click count.
* WebKitTestRunner/InjectedBundle/EventSendingController.h: Updated
for the changes above.
2011-03-31 Sergio Villar Senin <svillar@igalia.com>
Reviewed by Martin Robinson.
[GTK] implement LayoutTestController::setWillSendRequestReturnsNull
https://bugs.webkit.org/show_bug.cgi?id=57362
Do not generate DRT output if willSendRequestReturnsNull is set.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(willSendRequestCallback):
2011-03-31 Vamshikrishna.Yellenki <vamshi@motorola.com> and Alejandro G. Castro <alex@igalia.com>
Reviewed by Martin Robinson.
Implement MiniBrowser for Gtk port.
https://bugs.webkit.org/show_bug.cgi?id=48512
Initial implementation of the Gtk MiniBrowser.
* MiniBrowser/gtk/GNUmakefile.am: Added.
* MiniBrowser/gtk/main.c: Added.
(activateUriEntryCallback):
(destroyCallback):
(goBackCallback):
(goForwardCallback):
(createToolbar):
(createWebView):
(createWindow):
(argumentToURL):
(main):
2011-03-30 Dominic Cooney <dominicc@google.com>
Reviewed by Dimitri Glazkov.
Adds layoutTestController.shadowRoot accessor to Mac DRT.
https://bugs.webkit.org/show_bug.cgi?id=57415
* DumpRenderTree/LayoutTestController.cpp:
(shadowRootCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::shadowRoot):
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::shadowRoot):
2011-03-30 Matthew Delaney <mdelaney@apple.com>
Reviewed by Chris Marrin.
Update fast/canvas tests to avoid dumping the render tree when possible
https://bugs.webkit.org/show_bug.cgi?id=57493
* DumpRenderTree/mac/DumpRenderTree.mm: Make DRT aware of new default values for
accelerated drawing and accelerated drawing for canvas
2011-03-30 Adam Roben <aroben@apple.com>
Stop ignoring leaks in CGGradientCreateWithColorStops
Fixes <rdar://problem/7888547>.
Rubber-stamped by John Sullivan.
* Scripts/old-run-webkit-tests:
(countAndPrintLeaks): Removed some code to ignore those leaks.
2011-03-30 Timur Iskhodzhanov <timurrrr@google.com>
Reviewed by Alexey Proskuryakov.
Add some dynamic annotations to JavaScriptCore/wtf
https://bugs.webkit.org/show_bug.cgi?id=53747
By using these annotations we can improve the precision of finding
WebKit errors using dynamic analysis tools like ThreadSanitizer and Valgrind.
These annotations don't affect the compiled binaries unless USE(DYNAMIC_ANNOTATIONS) is "1".
These files don't add new functionality, so don't need extra tests.
* DumpRenderTree/ForwardingHeaders/wtf/DynamicAnnotations.h: Added.
2011-03-30 Steve Falkenburg <sfalken@apple.com>
Reviewed by Adam Roben.
Share most vsprops between Release and Production builds in releaseproduction.vsprops
https://bugs.webkit.org/show_bug.cgi?id=57508
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops:
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginRelease.vsprops:
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseCairoCFLite.vsprops:
* DumpRenderTree/win/DumpRenderTreeProduction.vsprops:
* DumpRenderTree/win/DumpRenderTreeRelease.vsprops:
* DumpRenderTree/win/DumpRenderTreeReleaseCairoCFLite.vsprops:
* DumpRenderTree/win/ImageDiffProduction.vsprops:
* DumpRenderTree/win/ImageDiffRelease.vsprops:
* DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops:
* FindSafari/FindSafariProduction.vsprops:
* FindSafari/FindSafariRelease.vsprops:
* FindSafari/FindSafariReleaseCairoCFLite.vsprops:
* FindSafari/FindSafariReleasePGO.vsprops:
* MiniBrowser/Configurations/MiniBrowserProduction.vsprops:
* MiniBrowser/Configurations/MiniBrowserRelease.vsprops:
* MiniBrowser/Configurations/MiniBrowserReleaseCairoCFLite.vsprops:
* TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops:
* TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops:
* WebKitAPITest/WebKitAPITestProduction.vsprops:
* WebKitAPITest/WebKitAPITestRelease.vsprops:
* WebKitAPITest/WebKitAPITestReleaseCairoCFLite.vsprops:
* WebKitLauncherWin/WebKitLauncherWinProduction.vsprops:
* WebKitLauncherWin/WebKitLauncherWinRelease.vsprops:
* WebKitLauncherWin/WebKitLauncherWinReleaseCairoCFLite.vsprops:
* WebKitTestRunner/win/InjectedBundleProduction.vsprops:
* WebKitTestRunner/win/InjectedBundleRelease.vsprops:
* WebKitTestRunner/win/InjectedBundleReleaseCairoCFLite.vsprops:
* WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops:
* WebKitTestRunner/win/WebKitTestRunnerRelease.vsprops:
* WebKitTestRunner/win/WebKitTestRunnerReleaseCairoCFLite.vsprops:
* WinLauncher/WinLauncherProduction.vsprops:
* WinLauncher/WinLauncherRelease.vsprops:
* WinLauncher/WinLauncherReleaseCairoCFLite.vsprops:
* record-memory-win/record-memory-winProduction.vsprops:
* record-memory-win/record-memory-winRelease.vsprops:
* record-memory-win/record-memory-winReleaseCairoCFLite.vsprops:
2011-03-30 Steve Falkenburg <sfalken@apple.com>
Reviewed by Adam Roben.
Update Windows production build logic for new production configurations
https://bugs.webkit.org/show_bug.cgi?id=57494
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops:
* DumpRenderTree/win/DumpRenderTreeProduction.vsprops:
* DumpRenderTree/win/ImageDiffProduction.vsprops:
* FindSafari/FindSafariProduction.vsprops:
* FindSafari/FindSafariReleasePGO.vsprops:
* MiniBrowser/Configurations/MiniBrowserProduction.vsprops:
* WebKitAPITest/WebKitAPITestProduction.vsprops:
* WebKitLauncherWin/WebKitLauncherWinProduction.vsprops:
* WebKitTestRunner/win/InjectedBundleProduction.vsprops:
* WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops:
* WinLauncher/WinLauncherProduction.vsprops:
* record-memory-win/record-memory-winProduction.vsprops:
2011-03-30 Robert Hogan <robert@webkit.org>
Reviewed by Antonio Gomes.
[Qt] Fix LoadHTMLStringItem::invoke() after r75966
Unskip http/tests/navigation/go-back-to-error-page.html
Also add the location of DumpRenderTreeSupportQt.h
to DRT's include paths.
https://bugs.webkit.org/show_bug.cgi?id=52614
* DumpRenderTree/qt/DumpRenderTree.pro:
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
* DumpRenderTree/qt/DumpRenderTreeQt.h:
* DumpRenderTree/qt/GCControllerQt.cpp:
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::queueLoadHTMLString):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/qt/PlainTextControllerQt.cpp:
* DumpRenderTree/qt/TextInputControllerQt.cpp:
* DumpRenderTree/qt/WorkQueueItemQt.cpp:
(LoadAlternateHTMLStringItem::invoke):
* DumpRenderTree/qt/WorkQueueItemQt.h:
(LoadAlternateHTMLStringItem::LoadAlternateHTMLStringItem):
* QtTestBrowser/QtTestBrowser.pro:
* QtTestBrowser/launcherwindow.h:
2011-03-30 Adam Barth <abarth@webkit.org>
Reviewed by Adam Roben.
SheriffBot rollouts take too long
https://bugs.webkit.org/show_bug.cgi?id=57498
We used to build before landing rollouts via the commit-queue to
prevent further breakage, but now that our individual commit-queue
machines are slower, building takes too long. I can't remember the
last time a rollout broke compile. It seems like just landing the
patch is the better trade-off.
* Scripts/webkitpy/tool/bot/commitqueuetask.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
2011-03-30 Adam Barth <abarth@webkit.org>
Remove platform/chromium-mac-snowleopard
https://bugs.webkit.org/show_bug.cgi?id=57486
This directory no longer exists, so we can remove it from the fallback
chain.
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2011-03-30 Steve Falkenburg <sfalken@apple.com>
Reviewed by Adam Roben.
Rename Windows configuration Release_LTCG to Production for clarity
https://bugs.webkit.org/show_bug.cgi?id=57465
* DumpRenderTree/DumpRenderTree.sln:
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops: Copied from Tools/DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops: Removed.
* DumpRenderTree/win/DumpRenderTree.vcproj:
* DumpRenderTree/win/DumpRenderTreeProduction.vsprops: Copied from Tools/DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops.
* DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops: Removed.
* DumpRenderTree/win/ImageDiff.vcproj:
* DumpRenderTree/win/ImageDiffProduction.vsprops: Copied from Tools/DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops.
* DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops: Removed.
* FindSafari/FindSafari.vcproj:
* FindSafari/FindSafariProduction.vsprops: Copied from Tools/FindSafari/FindSafariReleaseLTCG.vsprops.
* FindSafari/FindSafariReleaseLTCG.vsprops: Removed.
* MiniBrowser/Configurations/MiniBrowserProduction.vsprops: Copied from Tools/MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops.
* MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops: Removed.
* MiniBrowser/MiniBrowser.vcproj:
* Scripts/webkitdirs.pm:
* TestWebKitAPI/win/TestWebKitAPI.sln:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
* TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj:
* TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
* WebKitAPITest/WebKitAPITest.vcproj:
* WebKitAPITest/WebKitAPITestProduction.vsprops: Copied from Tools/WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops.
* WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops: Removed.
* WebKitLauncherWin/WebKitLauncherWin.vcproj:
* WebKitLauncherWin/WebKitLauncherWinProduction.vsprops: Copied from Tools/WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops.
* WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops: Removed.
* WebKitTestRunner/WebKitTestRunner.sln:
* WebKitTestRunner/win/InjectedBundle.vcproj:
* WebKitTestRunner/win/InjectedBundleGenerated.vcproj:
* WebKitTestRunner/win/InjectedBundleProduction.vsprops: Copied from Tools/WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops.
* WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops: Removed.
* WebKitTestRunner/win/WebKitTestRunner.vcproj:
* WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops: Copied from Tools/WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops.
* WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops: Removed.
* WinLauncher/WinLauncher.vcproj:
* WinLauncher/WinLauncherProduction.vsprops: Copied from Tools/WinLauncher/WinLauncherReleaseLTCG.vsprops.
* WinLauncher/WinLauncherReleaseLTCG.vsprops: Removed.
* record-memory-win/record-memory-win.vcproj:
* record-memory-win/record-memory-winProduction.vsprops: Copied from Tools/record-memory-win/record-memory-winReleaseLTCG.vsprops.
* record-memory-win/record-memory-winReleaseLTCG.vsprops: Removed.
2011-03-30 MORITA Hajime <morrita@google.com>
Reviewed by Dimitri Glazkov.
[Chromium] Expose the shadow DOM to DumpRenderTree JS tests.
https://bugs.webkit.org/show_bug.cgi?id=56573
Added LayoutTestController.shadowRoot() to Chromium DRT.
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::shadowRoot):
* DumpRenderTree/chromium/LayoutTestController.h:
2011-03-30 Yuta Kitamura <yutak@chromium.org>
Reviewed by Kent Tamura.
[Chromium] DumpRenderTree: Implement LayoutTestController::setPluginsEnabled
https://bugs.webkit.org/show_bug.cgi?id=57430
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setPluginsEnabled):
* DumpRenderTree/chromium/LayoutTestController.h:
2011-03-30 Maciej Stachowiak <mjs@apple.com>
Reviewed by Adam Barth.
make webkit-patch upload respect -d
https://bugs.webkit.org/show_bug.cgi?id=57425
* Scripts/webkitpy/common/checkout/scm.py:
* Scripts/webkitpy/tool/steps/commit.py:
2011-03-30 Maciej Stachowiak <mjs@apple.com>
Reviewed by Adam Barth.
Make "webkit-patch --dry-run --verbose land" log the SVN command it's going to use
https://bugs.webkit.org/show_bug.cgi?id=57429
* Scripts/webkitpy/common/checkout/scm.py:
2011-03-29 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r82383.
http://trac.webkit.org/changeset/82383
https://bugs.webkit.org/show_bug.cgi?id=57417
"nrwt isn't shutting down cleanly" (Requested by dpranke on
#webkit).
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-03-29 Kent Tamura <tkent@chromium.org>
Reviewed by Dimitri Glazkov.
Make validation message bubble testable
https://bugs.webkit.org/show_bug.cgi?id=57290
* DumpRenderTree/mac/DumpRenderTree.mm:
(createWebViewAndOffscreenWindow): Change a setting so that validation
bubbles isn't hidden automatically.
2011-03-29 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: use 'threads' on win instead of 'old-threads'
It's a bit unclear whether multiple threads or multiple
processes will be a better model on windows. This change will
test threads for now while we're still working out the bugs in
multiple-processes.
https://bugs.webkit.org/show_bug.cgi?id=57410
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-03-29 David Levin <levin@chromium.org>
Reviewed by Shinichiro Hamaji.
check-webkit-style confused by two ChangeLog entries in a row from same user
https://bugs.webkit.org/show_bug.cgi?id=57250
* Scripts/webkitpy/style/checker.py: Add the line should be checked function to ChangeLogChecker.
* Scripts/webkitpy/style/checker_unittest.py: Fix test due to that new function.
* Scripts/webkitpy/style/checkers/changelog.py: Made this code aware of what lines were being checked.
It basically assumes only one ChangeLog entry is being processed because that is the standard case and
checking more than that would be very messey.
* Scripts/webkitpy/style/checkers/changelog_unittest.py: Add testing to catch the broken case.
* Scripts/webkitpy/style/error_handlers.py: Added should_line_be_checked.
2011-03-29 Kent Tamura <tkent@chromium.org>
Reviewed by Dimitri Glazkov.
[Mac] Enable interactive-validation tests on Mac DRT
https://bugs.webkit.org/show_bug.cgi?id=57308
* DumpRenderTree/mac/DumpRenderTree.mm:
(createWebViewAndOffscreenWindow): Enable the form interactive validation feature.
2011-03-29 Chang Shu <cshu@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
WebKitTestRunner needs layoutTestController.setAllowUniversalAccessFromFileURLs
https://bugs.webkit.org/show_bug.cgi?id=42692
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::setAllowUniversalAccessFromFileURLs):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2011-03-24 Martin Robinson <mrobinson@igalia.com>
Reviewed by Xan Lopez.
[GTK] [DRT] GtkScrolledWindow adds three pixels of padding between the WebView and the scrollbar
https://bugs.webkit.org/show_bug.cgi?id=57067
Eliminate extra spacing between the DRT GtkScrolledWindow and the DRT WebView. This
will allow WebKit1GTK+ and WebKit2GTK+ to share test results. Followup commits will
update DRT and pixel results as hundreds of results will need to be updated.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(setDefaultsToConsistentStateValuesForTesting): Eliminate spacing on the GtkScrolledWindow
by overriding the theme RC/CSS files.
2011-03-29 Steve Falkenburg <sfalken@apple.com>
Reviewed by Adam Roben.
Use per-configuration vsprops in tools projects to avoid WebKitVSPropsRedirectionDir removal by MSVC IDE
https://bugs.webkit.org/show_bug.cgi?id=57394
Visual Studio's IDE was removing instances of $(WebKitVSPropsRedirectionDir) from
InheritedPropertySheet rules in our vcproj files when the vcproj was edited from within
the IDE. To avoid this, add a separate vsprops file for each project configuration that
contains the required inherited property sheets.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebug.vsprops: Added.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebugAll.vsprops: Added.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebugCairoCFLite.vsprops: Added.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginRelease.vsprops: Added.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseCairoCFLite.vsprops: Added.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops: Added.
* DumpRenderTree/win/DumpRenderTree.vcproj:
* DumpRenderTree/win/DumpRenderTreeDebug.vsprops: Added.
* DumpRenderTree/win/DumpRenderTreeDebugAll.vsprops: Added.
* DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Added.
* DumpRenderTree/win/DumpRenderTreeRelease.vsprops: Added.
* DumpRenderTree/win/DumpRenderTreeReleaseCairoCFLite.vsprops: Added.
* DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops: Added.
* DumpRenderTree/win/ImageDiff.vcproj:
* DumpRenderTree/win/ImageDiffDebug.vsprops: Added.
* DumpRenderTree/win/ImageDiffDebugAll.vsprops: Added.
* DumpRenderTree/win/ImageDiffDebugCairoCFLite.vsprops: Added.
* DumpRenderTree/win/ImageDiffRelease.vsprops: Added.
* DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops: Added.
* DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops: Added.
* FindSafari/FindSafari.vcproj:
* FindSafari/FindSafariDebug.vsprops: Added.
* FindSafari/FindSafariDebugAll.vsprops: Added.
* FindSafari/FindSafariDebugCairoCFLite.vsprops: Added.
* FindSafari/FindSafariRelease.vsprops: Added.
* FindSafari/FindSafariReleaseCairoCFLite.vsprops: Added.
* FindSafari/FindSafariReleaseLTCG.vsprops: Added.
* FindSafari/FindSafariReleasePGO.vsprops: Added.
* MiniBrowser/Configurations/MiniBrowserDebug.vsprops: Added.
* MiniBrowser/Configurations/MiniBrowserDebugAll.vsprops: Added.
* MiniBrowser/Configurations/MiniBrowserDebugCairoCFLite.vsprops: Added.
* MiniBrowser/Configurations/MiniBrowserRelease.vsprops: Added.
* MiniBrowser/Configurations/MiniBrowserReleaseCairoCFLite.vsprops: Added.
* MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops: Added.
* MiniBrowser/MiniBrowser.vcproj:
* WebKitAPITest/WebKitAPITest.vcproj:
* WebKitAPITest/WebKitAPITestDebug.vsprops: Added.
* WebKitAPITest/WebKitAPITestDebugAll.vsprops: Added.
* WebKitAPITest/WebKitAPITestDebugCairoCFLite.vsprops: Added.
* WebKitAPITest/WebKitAPITestRelease.vsprops: Added.
* WebKitAPITest/WebKitAPITestReleaseCairoCFLite.vsprops: Added.
* WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops: Added.
* WebKitLauncherWin/WebKitLauncherWin.vcproj:
* WebKitLauncherWin/WebKitLauncherWinDebug.vsprops: Added.
* WebKitLauncherWin/WebKitLauncherWinDebugAll.vsprops: Added.
* WebKitLauncherWin/WebKitLauncherWinDebugCairoCFLite.vsprops: Added.
* WebKitLauncherWin/WebKitLauncherWinRelease.vsprops: Added.
* WebKitLauncherWin/WebKitLauncherWinReleaseCairoCFLite.vsprops: Added.
* WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops: Added.
* WebKitTestRunner/win/InjectedBundle.vcproj:
* WebKitTestRunner/win/InjectedBundleDebug.vsprops: Added.
* WebKitTestRunner/win/InjectedBundleDebugAll.vsprops: Added.
* WebKitTestRunner/win/InjectedBundleDebugCairoCFLite.vsprops: Added.
* WebKitTestRunner/win/InjectedBundleRelease.vsprops: Added.
* WebKitTestRunner/win/InjectedBundleReleaseCairoCFLite.vsprops: Added.
* WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops: Added.
* WebKitTestRunner/win/WebKitTestRunner.vcproj:
* WebKitTestRunner/win/WebKitTestRunnerDebug.vsprops: Added.
* WebKitTestRunner/win/WebKitTestRunnerDebugAll.vsprops: Added.
* WebKitTestRunner/win/WebKitTestRunnerDebugCairoCFLite.vsprops: Added.
* WebKitTestRunner/win/WebKitTestRunnerRelease.vsprops: Added.
* WebKitTestRunner/win/WebKitTestRunnerReleaseCairoCFLite.vsprops: Added.
* WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops: Added.
* WinLauncher/WinLauncher.vcproj:
* WinLauncher/WinLauncherDebug.vsprops: Added.
* WinLauncher/WinLauncherDebugAll.vsprops: Added.
* WinLauncher/WinLauncherDebugCairoCFLite.vsprops: Added.
* WinLauncher/WinLauncherRelease.vsprops: Added.
* WinLauncher/WinLauncherReleaseCairoCFLite.vsprops: Added.
* WinLauncher/WinLauncherReleaseLTCG.vsprops: Added.
* record-memory-win/record-memory-win.vcproj:
* record-memory-win/record-memory-winDebug.vsprops: Added.
* record-memory-win/record-memory-winDebugAll.vsprops: Added.
* record-memory-win/record-memory-winDebugCairoCFLite.vsprops: Added.
* record-memory-win/record-memory-winRelease.vsprops: Added.
* record-memory-win/record-memory-winReleaseCairoCFLite.vsprops: Added.
* record-memory-win/record-memory-winReleaseLTCG.vsprops: Added.
2011-03-29 Adam Barth <abarth@webkit.org>
Reviewed by Darin Adler.
test_runner2.py can crash due to undefined variable
https://bugs.webkit.org/show_bug.cgi?id=57356
Replace reference to undefined variable with a variable that's actually
defined. AFAICT, there's no way to test this code.
* Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
2011-03-29 Tony Chang <tony@chromium.org>
Reviewed by Eric Seidel.
cleanup rebaseline-chromium-webkit-tests
https://bugs.webkit.org/show_bug.cgi?id=57375
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
- Split _extract_and_add_new_baselines into 3 functions
- Reduce the amount of line wrapping (the file already has lines
over 80 col, so may as well try to make the file consistent)
- Remove unnecessary ()s
2011-03-29 Brent Fulgham <bfulgham@webkit.org>
Reviewed by Adam Roben.
Make WinCairo a core builder.
https://bugs.webkit.org/show_bug.cgi?id=57373
* Scripts/webkitpy/common/net/buildbot/buildbot.py: Add WinCairo.
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py: Add WinCairo.
2011-03-29 Timothy Hatcher <timothy@apple.com>
Update update-webkit-localizable-strings to understand that WebKit and
WebKit2 strings go in WebCore.
Also make extract-localizable-strings optionally use an exclude file. No individual
warnings are produced about unlocalized strings when there is no exclude file. This
is needed for WebCore, since there are too many bare strings.
https://webkit.org/b/57354
Reviewed by Sam Weinig.
* Scripts/extract-localizable-strings: Support "-" to mean no exclude file.
* Scripts/update-webkit-localizable-strings: Update paths to scan WebCore and WebKit2.
2011-03-29 Darin Adler <darin@apple.com>
Reviewed by Adam Roben.
WebKit2 bundle page needs to offer generated file hooks
https://bugs.webkit.org/show_bug.cgi?id=57279
(Land missing part of the original patch after r82289.)
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage): Added new callbacks.
2011-03-29 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[chromium] NRWT should be able to read checksums from png comments
https://bugs.webkit.org/show_bug.cgi?id=57280
* Scripts/read-checksum-from-png: Added. Utility script to read the checksum
from a file.
* Scripts/webkitpy/common/system/filesystem.py: Add open_binary_file_for_reading
* Scripts/webkitpy/common/system/filesystem_mock.py:
* Scripts/webkitpy/layout_tests/port/base.py: When loading a checksum,
if the -checksum.txt file doesn't exist, try looking in the png.
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/read_checksum_from_png.py: Added. Scan the first
2k for a png comment with the checksum
* Scripts/webkitpy/layout_tests/read_checksum_from_png_unittest.py: Added.
2011-03-29 Philippe Normand <pnormand@igalia.com>
Rubber-stamped by Gustavo Noronha Silva.
[GTK] http/tests/uri/username-with-no-hostname.html fails
https://bugs.webkit.org/show_bug.cgi?id=57325
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(willSendRequestCallback): URI host value check done case-insensitively.
2011-03-29 Philippe Normand <pnormand@igalia.com>
Reviewed by Martin Robinson.
[GTK] http/tests/uri/username-with-no-hostname.html fails
https://bugs.webkit.org/show_bug.cgi?id=57325
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(willSendRequestCallback): Check URI before submitting the request
to WebCore.
2011-03-29 Noel Gordon <noel.gordon@gmail.com>
Reviewed by Ojan Vafai.
[chromium] DRT EventSender: remove identity from dragTargetDragEnter() calls
https://bugs.webkit.org/show_bug.cgi?id=57303
* DumpRenderTree/chromium/EventSender.cpp:
(EventSender::doDragDrop):
(EventSender::beginDragWithFiles):
2011-03-29 Zoltan Horvath <zoltan@webkit.org>
Reviewed by Andreas Kling.
[Qt] Add -print-loaded-urls option to Qt's MiniBrowser and to QtTestBrowser
https://bugs.webkit.org/show_bug.cgi?id=57314
Provide possibility to print loaded urls to the standard output.
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):
(BrowserWindow::printURL):
* MiniBrowser/qt/BrowserWindow.h:
* MiniBrowser/qt/MiniBrowserApplication.cpp:
(MiniBrowserApplication::handleUserOptions):
* MiniBrowser/qt/MiniBrowserApplication.h:
(WindowOptions::WindowOptions):
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::initializeView):
(LauncherWindow::printURL):
(LauncherWindow::cloneWindow):
* QtTestBrowser/launcherwindow.h:
(WindowOptions::WindowOptions):
* QtTestBrowser/main.cpp:
(LauncherApplication::handleUserOptions):
2011-03-29 Gabor Loki <loki@webkit.org>
Rubber-stamped by Csaba Osztrogonác.
[Qt] Teach build-jsc how to build JavaScriptCore on Qt
https://bugs.webkit.org/show_bug.cgi?id=56918
* Scripts/build-jsc: Remove duplicated --qt options to avoid passing them to qmake.
2011-03-28 Gabor Loki <loki@webkit.org>
Reviewed by Csaba Osztrogonác.
[Qt] Teach build-jsc how to build JavaScriptCore on Qt
https://bugs.webkit.org/show_bug.cgi?id=56918
* Scripts/build-jsc:
* Scripts/webkitdirs.pm:
2011-03-28 Andrew Foster <andrewf@chromium.org>
Reviewed by Ojan Vafai.
webkit-patch should be more intelligent about whether a bug applies to a patch
Create a new function, parse_bug_id_from_changelog() which determines
that a bug is related to a patch by parsing the output generated from
prepare-ChangeLog, rather than arbitrarily matching a URL to a bug in
the description of a patch.
https://bugs.webkit.org/show_bug.cgi?id=56989
* Scripts/webkitpy/common/checkout/api.py:
* Scripts/webkitpy/common/checkout/changelog.py:
* Scripts/webkitpy/common/net/bugzilla/__init__.py:
* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
* Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
* Scripts/webkitpy/style/checkers/changelog.py:
* Scripts/webkitpy/tool/commands/upload.py:
2011-03-28 Maciej Stachowiak <mjs@apple.com>
Revert accidentally commited change.
* DumpRenderTree/mac/ResourceLoadDelegate.mm:
(-[ResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
2011-03-28 Tony Chang <tony@chromium.org>
Reviewed by Adam Barth.
[chromium] have DRT write the png checksum into a png comment
https://bugs.webkit.org/show_bug.cgi?id=57255
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::dumpImage):
2011-03-28 Adele Peterson <adele@apple.com>
Removing Qt stub, since it doesn't use the same header as the others.
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2011-03-28 Adele Peterson <adele@apple.com>
Added a few more stubs.
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::hasGrammarMarker):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::hasGrammarMarker):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::hasGrammarMarker):
2011-03-28 Adele Peterson <adele@apple.com>
Build fix.
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::hasGrammarMarker):
2011-03-28 Adele Peterson <adele@apple.com>
Reviewed by Eric Seidel.
Testing support for <rdar://problem/9112694> REGRESSION (r79411): "Check grammar with spelling" context menu doesn't check as you type
https://bugs.webkit.org/show_bug.cgi?id=57173
* DumpRenderTree/LayoutTestController.cpp:
(hasGrammarMarkerCallback): Added.
(LayoutTestController::staticFunctions): Added case for hasGrammarMarker.
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/mac/DumpRenderTree.mm: (createWebViewAndOffscreenWindow): Call setGrammarCheckingEnabled.
* DumpRenderTree/mac/LayoutTestControllerMac.mm: (LayoutTestController::hasGrammarMarker): Added. Call new hasGrammarMarker method.
2011-03-28 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Benjamin Poulain.
[Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
https://bugs.webkit.org/show_bug.cgi?id=57087
Trivial typo fix after r82082.
* WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
(WTR::activateFonts):
2011-03-28 Alexis Menard <alexis.menard@openbossa.org>
Reviewed by Andreas Kling.
Add myself to the committers list.
* Scripts/webkitpy/common/config/committers.py:
2011-03-28 Simon Fraser <simon.fraser@apple.com>
Reviewed by Sam Weinig.
WebKit2 unfortunately uses code from WebKit, so MiniBrowser needs
to link against WebKit.framework for now.
* MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
2011-03-28 Benjamin Poulain <benjamin.poulain@nokia.com>
Reviewed by Andreas Kling.
[Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
https://bugs.webkit.org/show_bug.cgi?id=57087
Disable QT_ASCII_CAST_WARNINGS for applications.
* DumpRenderTree/qt/DumpRenderTree.pro:
* DumpRenderTree/qt/ImageDiff.pro:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* MiniBrowser/qt/MiniBrowser.pro:
* QtTestBrowser/QtTestBrowser.pro:
* WebKitTestRunner/qt/WebKitTestRunner.pro:
2011-03-28 Andreas Kling <kling@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt] QtTestBrowser doesn't reset its window title when opening about:blank
https://bugs.webkit.org/show_bug.cgi?id=57224
* QtTestBrowser/mainwindow.cpp:
(MainWindow::buildUI):
(MainWindow::onTitleChanged):
* QtTestBrowser/mainwindow.h:
2011-03-27 Jer Noble <jer.noble@apple.com>
Reviewed by Maciej Stachowiak.
Full Screen: disable keyboard access by default
https://bugs.webkit.org/show_bug.cgi?id=56684
Accept the withKeyboard parameter to supportsFullScreenForElement.
* DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate webView:supportsFullScreenForElement:withKeyboard:]):
2011-03-27 Patrick Gansterer <paroga@webkit.org>
Reviewed by David Levin.
check-webkit-style should check ChangeLog for a valid bug number
https://bugs.webkit.org/show_bug.cgi?id=57184
* Scripts/webkitpy/style/checker.py:
* Scripts/webkitpy/style/checker_unittest.py:
* Scripts/webkitpy/style/checkers/changelog.py: Added.
* Scripts/webkitpy/style/checkers/changelog_unittest.py: Added.
2011-03-27 Benjamin Poulain <benjamin.poulain@nokia.com>
Reviewed by Andreas Kling.
[Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
https://bugs.webkit.org/show_bug.cgi?id=57087
Use explicit conversion for string to avoid depending on the default codec
installed by the user code.
* WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
(WTR::activateFonts):
2011-03-26 Maciej Stachowiak <mjs@apple.com>
Revert inadvertantly committed changes.
* Scripts/old-run-webkit-tests:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-03-26 Andreas Kling <kling@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Show page icons (favicons) in QtTestBrowser location bar.
https://bugs.webkit.org/show_bug.cgi?id=57162
* QtTestBrowser/QtTestBrowser.qrc:
* QtTestBrowser/favicon.png: Added.
* QtTestBrowser/locationedit.cpp:
(defaultPageIcon):
(LocationEdit::LocationEdit):
(LocationEdit::setPageIcon):
(LocationEdit::resizeEvent):
(LocationEdit::updateInternalGeometry):
* QtTestBrowser/locationedit.h:
* QtTestBrowser/mainwindow.cpp:
(MainWindow::buildUI):
(MainWindow::onIconChanged):
(MainWindow::onLoadStarted):
* QtTestBrowser/mainwindow.h:
2011-03-23 Martin Robinson <mrobinson@igalia.com>
Reviewed by Xan Lopez.
[GTK] Force DumpRenderTree to use 96 DPI
https://bugs.webkit.org/show_bug.cgi?id=56866
Make sure that GTK+ is set to use 96 DPI when running tests. This ensures
broader compatibility with WebKit2 and makes setting font sizes more straightforward.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(initializeGtkFontSettings): Set the screen resolution and GTK+ xft property to 96 DPI.
(resetDefaultsToConsistentValues):Remove the call which repeatedly set the GDK screen resolution.
(setDefaultsToConsistentStateValuesForTesting): Update the font sizes to reflect the new DPI.
2011-03-25 Andy Estes <aestes@apple.com>
Reviewed by Adele Peterson.
REGRESSION (r70748): latest nightly builds kills AC_QuickTime.js
https://bugs.webkit.org/show_bug.cgi?id=49016
TestNetscapePlugIn needs to register support for an image MIME type so
we can test that the embed tag prefers plug-ins to render image MIME
types.
* DumpRenderTree/TestNetscapePlugIn/mac/Info.plist: Register image/png.
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(NP_GetMIMEDescription): Ditto.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.rc: Ditto.
2011-03-25 Jessie Berlin <jberlin@apple.com>
Reviewed by Sam Weinig.
WebKit2: Need to be able to set and get the Cookie Storage Policy.
https://bugs.webkit.org/show_bug.cgi?id=50780
Add a test for getting and setting the HTTP Cookie Accept Policy in WebKit2.
The test is named "CookieManager" so that we can eventually exercise more of the
CookieManager functionality in the test.
* TestWebKitAPI/Tests/WebKit2/CookieManager.cpp: Added.
(TestWebKitAPI::didGetTestHTTPCookieAcceptPolicy):
Assert that the policy returned is equal to the policy set in
didGetUserHTTPCookieAcceptPolicy, and then restore the user's policy.
(TestWebKitAPI::didGetUserHTTPCookieAcceptPolicy):
Set the policy to something different than the user's policy and get the policy again.
(TestWebKitAPI::didFinishLoadForFrame):
Get the user's policy so that it can be restored at the end of the test.
(TestWebKitAPI::TEST):
Load about:blank so that the Web Process gets instatiated (needed to get the
CookieManager).
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
Add CookieManager.cpp.
2011-03-24 Kevin Ollivier <kevino@theolliviers.com>
[wx] Fix the build so that gcc-4.1 and up can be used to build on SnowLeopard.
* wx/build/settings.py:
2011-03-24 Ilya Sherman <isherman@chromium.org>
Reviewed by Adam Roben.
Implement layoutTestController.setAutoFilled in DRT on Windows
https://bugs.webkit.org/show_bug.cgi?id=56828
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::setAutofilled): Implemented.
2011-03-24 Brent Fulgham <bfulgham@webkit.org>
Unreviewed build correction.
Correct vsprops file for Debug variant of the CFLite build.
Disable a CFNETWORK-specific test when building for CFLite.
* TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
2011-03-24 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
fix regressions introduced in 81908 - there were a couple of
code paths that only fired under python 2.5 that I missed.
Also change a couple of 'python' references to sys.executable
in order to handle runing test-webkitpy with a binary other than
something called 'python'.
https://bugs.webkit.org/show_bug.cgi?id=57063
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
* Scripts/webkitpy/test/cat.py:
* Scripts/webkitpy/test/echo.py:
2011-03-24 Brent Fulgham <bfulgham@webkit.org>
Unreviewed build correction.
Add a stub implementation of the TestInvocation dumping logic.
Update project files to reflect new file.
* WebKitTestRunner/cairo: Added.
* WebKitTestRunner/cairo/TestInvocationCairo.cpp: Added.
(WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
* WebKitTestRunner/win/WebKitTestRunner.vcproj:
2011-03-24 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
new-run-webkit-tests: clean up worker model defaults. This
changes the default behavior to 'threads' instead of
'old-threads', and clarifies that chromium-win-* and
chromium-mac-leopard are exceptions.
https://bugs.webkit.org/show_bug.cgi?id=56971
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-03-23 Leandro Gracia Gil <leandrogracia@chromium.org>
Reviewed by Darin Adler.
Media Stream API: fix the order of the options in build-webkit.
https://bugs.webkit.org/show_bug.cgi?id=56949
Fix the order of the options in build-webkit.
* Scripts/build-webkit:
2011-03-23 Adam Klein <adamk@chromium.org>
Reviewed by Mihai Parparita.
Add an option to new-run-webkit-httpd to allow overriding the location of LayoutTests directory
https://bugs.webkit.org/show_bug.cgi?id=56884
This option will be utilized by Chromium's ui_tests to allow
referencing js-test-resources from a layout test over HTTP.
* Scripts/new-run-webkit-httpd:
Added --layout_tests_dir option.
* Scripts/webkitpy/layout_tests/port/http_server.py:
Plumbed through as layout_tests_dir param.
2011-03-23 Dirk Pranke <dpranke@chromium.org>
Reviewed by Mihai Parparita.
new-run-webkit-tests: run tests in ascending alphabetical order per
dir. This is a second attempt at the fix; the first attempt, in
r81597, broke the behavior on the old-inline and old-threads
worker models.
https://bugs.webkit.org/show_bug.cgi?id=56760
* Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-03-23 Mario Sanchez Prada <msanchez@igalia.com>
Reviewed by Martin Robinson.
[GTK] GTK's DRT to ouput detail's string for the 'property-change' signal
https://bugs.webkit.org/show_bug.cgi?id=56953
Print the string for the detail of 'property-change' signal.
* DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
(axObjectEventListener): Print the detail string.
2011-03-23 Yury Semikhatsky <yurys@chromium.org>
Reviewed by Pavel Feldman.
[V8] Web Inspector: compile DebuggerScript.js into DebuggerScriptSource.h
https://bugs.webkit.org/show_bug.cgi?id=56843
* DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
* DumpRenderTree/chromium/DRTDevToolsAgent.h:
2011-03-22 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Teach update-webkit about gyp
https://bugs.webkit.org/show_bug.cgi?id=56881
This makes it possible for someone who builds from Xcode
to just run update-webkit --gyp and have the project files
generated after update automatically.
I also added a stand-alone "generate-project-files" script
wrapping Source/gyp/configure. This makes running gyp simpler
since most people will not have gyp or Source/gyp in their path.
* Scripts/build-webkit:
* Scripts/generate-project-files: Added.
* Scripts/update-webkit:
2011-03-22 Anton D'Auria <adauria@apple.com>
Reviewed by Alexey Proskuryakov.
Add +[WebApplicationCache getOriginsWithCache]
https://bugs.webkit.org/show_bug.cgi?id=56722
* DumpRenderTree/LayoutTestController.cpp:
(originsWithApplicationCacheCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
(LayoutTestController::originsWithApplicationCache):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
(LayoutTestController::originsWithApplicationCache):
* DumpRenderTree/mac/LayoutTestControllerMac.mm: Pass array of origin's database identifiers.
(originsArrayToJS): Helper function for converting array of WebSecurityOrigins to a JS array of origin identifiers.
(LayoutTestController::originsWithApplicationCache):
(LayoutTestController::originsWithLocalStorage):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
(LayoutTestController::originsWithApplicationCache):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
(LayoutTestController::originsWithApplicationCache):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp: Added stub.
(LayoutTestController::originsWithApplicationCache):
2011-03-22 Eric Seidel <eric@webkit.org>
Reviewed by Adam Barth.
Add support to build-webkit for building with gyp-generated project files
https://bugs.webkit.org/show_bug.cgi?id=56877
build-webkit is a cesspool. This change just hacks in --gyp support for Mac.
Eventually we'll want to support gyp-generated projects in a more general manner.
* Scripts/build-webkit:
2011-03-22 David Kilzer <ddkilzer@apple.com>
<http://webkit.org/b/56781> Add --dsym switch to enable dsym generation when building with Xcode
Reviewed by Joseph Pecoraro.
* Scripts/build-webkit: Updated usage statement to include
--dsym.
* Scripts/webkitdirs.pm: Added $generateDsym variable.
(generateDsym): Added. Call determineGenerateDsym() and
return $generateDsym.
(determineGenerateDsym): Added. Parse @ARGV for --dsym and set
$generateDsym.
(argumentsForXcode): Added. Returns additional arguments for
xcodebuild based on command-line switches.
(XcodeOptions): Updated to include argumentsForXcode() in the
array returned.
2011-03-22 Adam Roben <aroben@apple.com>
Make Leopard Debug bots and Windows XP Debug (Tests) core builders
These bots are consistently green again.
Fixes <http://webkit.org/b/56830> Leopard Debug bots and Windows XP Debug (Tests) should be
core builders
Rubber-stamped by Antti Koivisto.
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
(BuildBot.__init__): Loosened the regexps for Leopard and Windows.
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
(BuildBotTest.test_builder_name_regexps): Updated expectations.
2011-03-21 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r81597.
http://trac.webkit.org/changeset/81597
https://bugs.webkit.org/show_bug.cgi?id=56801
Change results in lots of unexpected flaky on the dashboard.
(Requested by pfeldman on #webkit).
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-03-21 Sam Weinig <sam@webkit.org>
Fix failing tests on the WebKit2 bots.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::resetStateToConsistentValues):
2011-03-21 Daniel Sievers <sievers@google.com>
Reviewed by Simon Fraser.
[Chromium] Make RenderAsTextBehavior and LayerTreeAsTextBehavior tweakable from the DumpRenderTree commandline
https://bugs.webkit.org/show_bug.cgi?id=56139
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::layerTreeAsText):
* DumpRenderTree/chromium/LayoutTestController.h:
(LayoutTestController::setShowDebugLayerTree):
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::runFileTest):
(TestShell::dump):
* DumpRenderTree/chromium/TestShell.h:
(TestParams::TestParams):
2011-03-21 Adam Roben <aroben@apple.com>
Try again to get buildbot to show run-javascriptcore-tests's actual.html as an HTML file
Fixes <http://webkit.org/b/56746> build.webkit.org shows run-javascriptcore-tests's HTML
output as plain text
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(RunJavaScriptCoreTests): List actual.html as one of our log files so it will get uploaded
to the master.
(RunJavaScriptCoreTests.commandComplete): Reading the actual.html file here won't work,
since we're running on the build master, not the slave. Instead, turn the raw source of
actual.html which the build slave uploaded into an HTML log.
2011-03-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Mihai Parparita.
new-run-webkit-tests: run tests in ascending alphabetical order per dir
https://bugs.webkit.org/show_bug.cgi?id=56760
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
* Scripts/webkitpy/layout_tests/port/test.py:
2011-03-21 Leandro Gracia Gil <leandrogracia@chromium.org>
Reviewed by Steve Block.
Media Stream API patch 0: adding compilation guards.
https://bugs.webkit.org/show_bug.cgi?id=56458
Adding options to enable the media stream API feature in Chromium.
* Scripts/build-webkit:
2011-03-21 Qi Zhang <qi.2.zhang@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] Add a command line option to capture stdout and stderr for DumpRenderTree
https://bugs.webkit.org/show_bug.cgi?id=56323
Using freopen to redirect STDOUT and STDERR when DumpRenderTree command line provide
"--stdout" or "--stderr" option.
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::~DumpRenderTree):
(WebCore::DumpRenderTree::processArgsLine):
(WebCore::DumpRenderTree::loadNextTestInStandAloneMode):
* DumpRenderTree/qt/DumpRenderTreeQt.h:
(WebCore::DumpRenderTree::setRedirectOutputFileName):
(WebCore::DumpRenderTree::setRedirectErrorFileName):
* DumpRenderTree/qt/main.cpp:
(isOption):
(takeOptionValue):
(printUsage):
(main):
2011-03-21 Adam Roben <aroben@apple.com>
Fix exceptions on the buildbot due to my last change
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(RunJavaScriptCoreTests.commandComplete): Catch exceptions due to actual.html not existing.
2011-03-21 Adam Roben <aroben@apple.com>
Teach buildbot to treat run-javascriptcore-tests's actual.html as an HTML file
Fixes <http://webkit.org/b/56746> build.webkit.org shows run-javascriptcore-tests's HTML
output as plain text
Reviewed by Darin Adler.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(RunJavaScriptCoreTests): Removed the logfiles variable, which can only handle plaintext
logs.
(RunJavaScriptCoreTests.commandComplete): Use the addHTMLLog method to upload actual.html to
the build master. This will correctly treat it as HTML.
2011-03-20 Bill Budge <bbudge@chromium.org>
Reviewed by Adam Barth.
Rename ThreadSafeShared to ThreadSafeRefCounted
https://bugs.webkit.org/show_bug.cgi?id=56714
No new tests. Exposes no new functionality.
* DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeRefCounted.h: Copied from DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeShared.h.
* DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeShared.h: Removed.
* Scripts/do-webcore-rename:
2011-03-19 Anton D'Auria <adauria@apple.com>
Reviewed by Dan Bernstein.
Add stub to LayoutTestControllerGtk.cpp to fix GTK build
https://bugs.webkit.org/show_bug.cgi?id=56719
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Adding stub.
(LayoutTestController::clearApplicationCacheForOrigin):
2011-03-19 Anton D'Auria <adauria@apple.com>
Reviewed by Alexey Proskuryakov.
ApplicationCacheGroup is not obsolete after being deleted via ApplicationCacheStorage::deleteEntriesForOrigin
https://bugs.webkit.org/show_bug.cgi?id=56415
* DumpRenderTree/LayoutTestController.cpp:
(clearApplicationCacheForOriginCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp: Adding stubs.
(LayoutTestController::LayoutTestController):
(LayoutTestController::clearAllApplicationCaches):
(LayoutTestController::clearApplicationCacheForOrigin):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/mac/LayoutTestControllerMac.mm: Adding stub.
(LayoutTestController::clearApplicationCacheForOrigin):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp: Adding stub.
(LayoutTestController::clearApplicationCacheForOrigin):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp: Adding stub.
(LayoutTestController::clearApplicationCacheForOrigin):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp: Adding stub.
(LayoutTestController::clearApplicationCacheForOrigin):
2011-03-18 Dan Bernstein <mitz@apple.com>
Revised build fix for r81135.
* WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2011-03-18 Adam Roben <aroben@apple.com>
Windows build fix
* WebKitTestRunner/config.h: Add a missing #endif.
2011-03-18 Andreas Kling <kling@webkit.org>
Unbreak Qt build after r81494.
* WebKitTestRunner/qt/TestInvocationQt.cpp:
(WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
2011-03-18 Sam Weinig <sam@webkit.org>
Fix windows build. This was really Adam Roben's fault. I swear!
* WebKitTestRunner/config.h:
2011-03-18 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
Pixel tests don't work in WebKitTestRunner on Windows
https://bugs.webkit.org/show_bug.cgi?id=56630
Cross-platformize (Mac and Windows CG) the pixel dumping code in WebKitTestRunner to use WebKit2
built-in snapshotting instead of using the window server on the Mac.
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::InjectedBundle):
(WTR::InjectedBundle::didReceiveMessage):
(WTR::InjectedBundle::beginTesting):
(WTR::InjectedBundle::done):
* WebKitTestRunner/InjectedBundle/InjectedBundle.h:
(WTR::InjectedBundle::setPixelResult):
(WTR::InjectedBundle::shouldDumpPixels):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::dump):
Take the snapshot in the bundle and send it to the UIProcess for dumping.
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::invoke):
(WTR::TestInvocation::dump):
(WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
* WebKitTestRunner/TestInvocation.h:
* WebKitTestRunner/cg: Added.
* WebKitTestRunner/cg/TestInvocationCG.cpp: Copied from WebKitTestRunner/mac/TestInvocationMac.mm.
(WTR::createCGContextFromImage):
(WTR::computeMD5HashStringForContext):
(WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
* WebKitTestRunner/mac/TestInvocationMac.mm: Removed.
Making pixel dumping code shared for CG ports. Also use the WTF/MD5 instead of platform specific code.
* WebKitTestRunner/win/TestInvocationWin.cpp:
(WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
Keep stub for Cairo build.
* WebKitTestRunner/Configurations/WebKitTestRunnerCoreGraphics.vsprops: Added.
Add CG configuration.
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
* WebKitTestRunner/win/WebKitTestRunner.vcproj:
Add new files.
2011-03-18 Adam Roben <aroben@apple.com>
Don't allow Leaks Viewer's vertical scrollbar to cover up long function names
Rubber-stamped by Simon Fraser.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
(.data-grid .data-container): Use overflow-y:auto instead of the Inspector's default
overflow-y:overlay so that the scrollbar doesn't overlay the content.
2011-03-18 Adam Roben <aroben@apple.com>
Don't ellipsize long or indented function names in Leaks Viewer
Fixes <http://webkit.org/b/56037> Leaks viewer needs horizontal
scrollbar
Reviewed by Simon Fraser.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
(.data-grid .data-container):
(.data-grid td > div, .data-grid th > div):
Allow the contents to extend to the right.
2011-03-18 Adam Roben <aroben@apple.com>
Fix some leaks in DRT seen on the leaks bot
Fixes <http://webkit.org/b/56638> addURLToRedirectCallback and
setWillSendRequestClearHeaderCallback in DRT are leaky
Reviewed by Alexey Proskuryakov.
* DumpRenderTree/ForwardingHeaders/wtf/OwnArrayPtr.h: Added.
* DumpRenderTree/LayoutTestController.cpp:
(addURLToRedirectCallback):
(setWillSendRequestClearHeaderCallback):
Use OwnArrayPtr to cause the strings to be deleted.
2011-03-18 Adam Roben <aroben@apple.com>
Fix a NSURLCredential leak seen on the leaks bot
Fixes <http://webkit.org/b/56637> -[SynchronousLoader
connection:didReceiveAuthenticationChallenge:] is leaky
Reviewed by John Sullivan.
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(-[SynchronousLoader connection:didReceiveAuthenticationChallenge:]): Use a RetainPtr to
cause the NSURLCredential to be released.
2011-03-18 John Knottenbelt <jknotten@chromium.org>
Unreviewed.
Remove unnecessary brackets around my email address.
* Scripts/webkitpy/common/config/committers.py:
2011-03-18 Adam Roben <aroben@apple.com>
Remove some files I accidentally added in r81454
* WebKitTestRunner/Configurations/WebKitTestRunnerCoreGraphics.vsprops: Removed.
* WebKitTestRunner/cg/TestInvocationCG.cpp: Removed.
2011-03-18 Adam Roben <aroben@apple.com>
Ignore an NSNumberFormatter leak on SnowLeopard
* Scripts/old-run-webkit-tests:
(countAndPrintLeaks): Ignore leaks in -[NSNumberFormatter
getObjectValue:forString:errorDescription:].
2011-03-18 Adam Roben <aroben@apple.com>
Fix a leak beneath LayoutTestController::authenticateSession seen on the bots
Rubber-stamped by Mark Rowe.
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::authenticateSession): Use a RetainPtr to cause the NSURLRequest to be
released.
2011-03-17 Adam Roben <aroben@apple.com>
Address a review comment I missed in r81445
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
(LeaksParserWorker.prototype._parseLeaks): Anchor the RegExp used to find the number of
leaked bytes to speed up searching on long lines.
2011-03-17 Adam Roben <aroben@apple.com>
Mark leaky builds orange, even if all regression tests passed
Fixes <http://webkit.org/b/56609> SnowLeopard Intel Leaks bot appears green on
build.webkit.org when all tests pass even though there are still leaks
Reviewed by Dan Bernstein.
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(RunWebKitLeakTests): Set the warnOnWarnings property to True so that the entire build will
be marked orange if this build step generates warnings.
2011-03-17 Adam Roben <aroben@apple.com>
Allow values in Leaks Viewer to be shown as percentages
Double-clicking on a column will toggle between showing percentages and showing real values.
Fixes <http://webkit.org/b/56628> Would like to view values as percentages of the total
number of leaked bytes in Leaks Viewer
Reviewed by Joe Pecoraro.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
(LeaksParserWorker.prototype._incorporateLeaks): Calculate the (never-shown) top-level
node's totalTime, which is used in calculating percentages.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js:
(monkeyPatchInspectorObjects): Only format values as bytes if we aren't supposed to be
showing them as percentages.
2011-03-17 Adam Roben <aroben@apple.com>
Show the number of leaked bytes, not just leaked allocations, in Leaks Viewer
The Self and Total columns now report the number of leaked bytes. The new Calls column shows
the number of leaked allocations. And the new Average column shows the average number of
leaked bytes per call.
Fixes <http://webkit.org/b/56344> Leaks Viewer should show how many bytes were leaked, not
just how many allocations
Reviewed by Joe Pecoraro.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
(LeaksParserWorker.prototype._parseLeaks): Parse the number of leaked bytes from the "Leak:"
line and store it with the leak stack.
(LeaksParserWorker.prototype._createNode): Initialize the averageTime and numberOfCalls
properties.
(LeaksParserWorker.prototype._incorporateLeaks): Updated to get the stack from the .stack
property, and changed to store the number of leaked bytes in .selfTime/.totalTime.
.numberOfCalls now holds the number of leaked allocations.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js:
(Preferences): Set samplingCPUProfiler to false so that the Average and Calls columns will
appear.
(monkeyPatchInspectorObjects): Format the values as bytes.
2011-03-17 Adam Roben <aroben@apple.com>
Make Leaks Viewer able to count leaks in builds that didn't fail any regression tests
Fixes <http://webkit.org/b/56626> Leaks Viewer omits recent builds which didn't fail any
regression tests
Reviewed by James Robinson.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
(RecentBuildsLoader.prototype.start): Get the number of leaks out of the "text" array, which
is always populated, even when all regression tests pass.
2011-03-10 Hayato Ito <hayato@chromium.org>
Reviewed by Tony Chang.
Add a check logic for reftests in case that a reftest is marked AS REBASELINE.
https://bugs.webkit.org/show_bug.cgi?id=56076
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-03-17 Adam Roben <aroben@apple.com>
Skip a compositing-sensitive test when accelerated compositing is disabled
Reviewed by Sam Weinig.
* Scripts/old-run-webkit-tests: Skip media/media-document-audio-repaint.html when
accelerated compositing is disabled.
2011-03-17 Adam Roben <aroben@apple.com>
Start using PlatformSpecificScheduler for the Mac builders
The current set of file/directory patterns should work for Mac builders. We may discover
that there are more files and directories we can exclude from the Mac builds, but they can
be added later.
Fixes <http://webkit.org/b/56593> Mac builders do lots of unnecessary builds
Reviewed by Mark Rowe.
* BuildSlaveSupport/build.webkit.org-config/config.json: Moved the Mac builders into two
PlatformSpecificSchedulers: one for Leopard, and one for SnowLeopard.
* Scripts/webkitpy/common/config/build.py:
(_should_file_trigger_build): Allow "mac-leopard" and "mac-snowleopard" platforms, and
cleaned up patterns a little bit.
* Scripts/webkitpy/common/config/build_unittest.py:
(ShouldBuildTest): Fixed some expected results.
(ShouldBuildTest.test_should_build): Start testing "mac-leopard" and "mac-snowleopard"
platforms.
2011-03-17 Jeff Miller <jeffm@apple.com>
Use a consistent set of file patterns in the svn:ignore property for all .xcodeproj directories, specifically:
*.mode*
*.pbxuser
*.perspective*
project.xcworkspace
xcuserdata
* DumpRenderTree/DumpRenderTree.xcodeproj: Modified property svn:ignore.
* MiniBrowser/MiniBrowser.xcodeproj: Modified property svn:ignore.
* TestWebKitAPI/TestWebKitAPI.xcodeproj: Modified property svn:ignore.
* WebKitLauncher/WebKitLauncher.xcodeproj: Modified property svn:ignore.
* WebKitTestRunner/WebKitTestRunner.xcodeproj: Modified property svn:ignore.
2011-03-17 Adam Roben <aroben@apple.com>
Make old-run-webkit-tests pick up mac-wk2 results when running in WebKit2 mode on Windows
Fixes <http://webkit.org/b/56563> Running WebKit2 tests on Windows should use results from
mac-wk2
Reviewed by Anders Carlsson.
* Scripts/old-run-webkit-tests:
(expectedDirectoryForTest): Clean up the code to search in extra Mac platform result
directories on Windows. Add mac-wk2 to the set of extra platforms we look in on Windows when
we're running in WebKit2 mode.
2011-03-17 Mikhail Naganov <mnaganov@chromium.org>
Reviewed by Yury Semikhatsky.
Web Inspector: Clean up Inspector strings.
https://bugs.webkit.org/show_bug.cgi?id=56557
Modify the script to deal correctly with "\xNN" escapes, and to
check strings in Chromium's DevTools.js.
* Scripts/check-inspector-strings:
2011-03-17 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Benjamin Poulain.
Update autoinstall to install mechanize 0.2.4
https://bugs.webkit.org/show_bug.cgi?id=56499
Newer versions of mechanize support HTTPS proxies, which is necessary
for people using webkit-patch behind firewalls.
* Scripts/webkitpy/common/system/autoinstall.py:
* Scripts/webkitpy/thirdparty/__init__.py:
2011-03-16 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
REGRESSION (r81084): media/invalid-media-url-crash.html and editing/execCommand/indent-paragraphs.html failing on WebKit2
https://bugs.webkit.org/show_bug.cgi?id=56541
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::willSendRequestForFrame):
Check for an empty host in addition to a null host. This bug was introduced
when porting off of the KURL code and onto the WKURLRef version.
2011-03-16 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
Add Lucas Forschler to the list of WebKit committers.
* Scripts/webkitpy/common/config/committers.py:
2011-03-16 Sam Weinig <sam@webkit.org>
Reviewed upon and dictated through Adam Roben.
Escape crashing function names in results.html. This fixes an issue
when a crash involves a type with templates.
* Scripts/old-run-webkit-tests:
2011-03-16 Daniel Sievers <sievers@google.com>
Reviewed by James Robinson.
Add setting to always force compositing mode
https://bugs.webkit.org/show_bug.cgi?id=56156
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
(TestShell::resetWebSettings):
* DumpRenderTree/chromium/TestShell.h:
(TestShell::setForceCompositingMode):
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::reset):
(WebPreferences::applyTo):
* DumpRenderTree/chromium/WebPreferences.h:
2011-03-15 Hayato Ito <hayato@chromium.org>
Reviewed by Ojan Vafai.
Ignore files that will be used by reftests in old-run-webkit-tests.
https://bugs.webkit.org/show_bug.cgi?id=55936
* Scripts/old-run-webkit-tests:
2011-03-15 Felipe Oliveira da Silva Netto <f.dachshund@gmail.com>
Reviewed by Adam Roben.
Function relativeScriptsDir doesn't return the correct relative
Scripts directory.
* Scripts/webkitdirs.pm:
(relativeScriptsDir): Ensure we use the directory that contains
webkitdirs.pm, not the one that contains the original script that
was invoked.
2011-03-15 Kevin Ollivier <kevino@theolliviers.com>
[wx] Build fixes for long linker commands on Win and method return error, plus a Mac
fix to link against the Security framework after recent WebKitSystemInterface changes.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::originsWithLocalStorage):
* wx/build/settings.py:
* wx/build/waf_extensions.py:
2011-03-15 Sam Weinig <sam@webkit.org>
Reviewed by Darin Adler.
Add script to convert a patch to a prettypatch and show in the default browser
https://bugs.webkit.org/show_bug.cgi?id=56416
This can be used either like
$ svn-create-patch | show-pretty-diff
or
$ show-pretty-diff patch.diff
* Scripts/show-pretty-diff: Added.
2011-03-15 Brian Weinstein <bweinstein@apple.com>
Rubber-stamped by Adam Roben.
Adding myself as a reviewer.
* Scripts/webkitpy/common/config/committers.py:
2011-03-15 Anders Carlsson <andersca@apple.com>
Try to fix the Windows build.
* DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp:
(NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
2011-03-15 Anders Carlsson <andersca@apple.com>
Reviewed by Sam Weinig.
Make sure that NP_Shutdown is always the last NPP function called
https://bugs.webkit.org/show_bug.cgi?id=56391
Add a test that times out if an NP_Deallocate object is called after NP_Shutdown.
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
(PluginTest::PluginTest):
(PluginTest::NP_Shutdown):
(PluginTest::registerNPShutdownFunction):
* DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
* DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp: Added.
(NPDeallocateCalledBeforeNPShutdown::NPDeallocateCalledBeforeNPShutdown):
(NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
(NPDeallocateCalledBeforeNPShutdown::ScriptableObject::hasProperty):
(NPDeallocateCalledBeforeNPShutdown::ScriptableObject::getProperty):
(NPDeallocateCalledBeforeNPShutdown::NPP_New):
(NPDeallocateCalledBeforeNPShutdown::NPP_GetValue):
(NPDeallocateCalledBeforeNPShutdown::shutdown):
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(NP_Shutdown):
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* GNUmakefile.am:
2011-03-15 Ilya Sherman <isherman@chromium.org>
Reviewed by Tony Chang.
Autofilled form elements are assigned fixed background color but not text color
https://bugs.webkit.org/show_bug.cgi?id=48382
* DumpRenderTree/LayoutTestController.cpp:
(setAutofilledCallback): Added.
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setAutofilled): Added.
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::setAutofilled): Added.
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::setAutofilled): Added.
(LayoutTestController::setValueForUser): Style tweak.
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::setAutofilled): Added (stub).
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::setAutofilled): Added (stub).
2011-03-15 Dan Bernstein <mitz@apple.com>
Build fix after r81135.
* TestWebKitAPI/Configurations/InjectedBundle.xcconfig:
2011-03-15 Dan Bernstein <mitz@apple.com>
Build fix after r81135.
* WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
* WebKitTestRunner/mac/TestControllerMac.mm:
* WebKitTestRunner/mac/TestInvocationMac.mm:
* WebKitTestRunner/mac/main.mm:
2011-03-15 Balazs Kelemen <kbalazs@webkit.org>
Reviewed by Adam Barth.
REGRESSION(80977): breaks Qt ARMv7 Linux Release bot
https://bugs.webkit.org/show_bug.cgi?id=56349
* QtTestBrowser/QtTestBrowser.pro: Link against fontconfig
on platforms where we use it. Copied from DumpRenderTree.pro.
2011-03-15 Dan Bernstein <mitz@apple.com>
Build fix after r81135.
* DumpRenderTree/cf/WebArchiveDumpSupport.cpp:
* DumpRenderTree/mac/WebArchiveDumpSupportMac.mm:
2011-03-15 Kevin Ollivier <kevino@theolliviers.com>
Reviewed by Darin Adler.
Introduce WTF_USE_EXPORT_MACROS, which will allow us to put shared library import/export
info into the headers rather than in export symbol definition files, but disable it on
all platforms initially so we can deal with port build issues one port at a time.
https://bugs.webkit.org/show_bug.cgi?id=27551
* DumpRenderTree/cg/ImageDiffCG.cpp:
* DumpRenderTree/chromium/config.h:
* DumpRenderTree/config.h:
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
* TestWebKitAPI/TestWebKitAPIPrefix.h:
* WebKitAPITest/HostWindow.cpp:
* WebKitAPITest/TestsController.cpp:
* WebKitAPITest/config.h: Added.
* WebKitAPITest/main.cpp:
* WebKitAPITest/tests/WebViewDestruction.cpp:
* WebKitTestRunner/InjectedBundle/Bindings/JSWrapper.cpp:
* WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
* WebKitTestRunner/InjectedBundle/GCController.cpp:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
* WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
* WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
* WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
* WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
* WebKitTestRunner/InjectedBundle/win/ActivateFonts.cpp:
* WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
* WebKitTestRunner/InjectedBundle/win/LayoutTestControllerWin.cpp:
* WebKitTestRunner/TestController.cpp:
* WebKitTestRunner/TestInvocation.cpp:
* WebKitTestRunner/config.h: Added.
* WebKitTestRunner/qt/PlatformWebViewQt.cpp:
* WebKitTestRunner/qt/TestControllerQt.cpp:
* WebKitTestRunner/qt/TestInvocationQt.cpp:
* WebKitTestRunner/qt/main.cpp:
* WebKitTestRunner/win/PlatformWebViewWin.cpp:
* WebKitTestRunner/win/TestControllerWin.cpp:
* WebKitTestRunner/win/TestInvocationWin.cpp:
* WebKitTestRunner/win/main.cpp:
* wx/build/settings.py:
2011-03-15 Hayato Ito <hayato@chromium.org>
Reviewed by Shinichiro Hamaji.
[NRWT] Add support for reftests to new-run-webkit-tests.
https://bugs.webkit.org/show_bug.cgi?id=55457
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/dryrun.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/port/test_files.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-03-15 Alejandro G. Castro <alex@igalia.com>
Reviewed by Martin Robinson.
[GTK] Fix compilation warning after r80892
https://bugs.webkit.org/show_bug.cgi?id=56302
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::originsWithLocalStorage):
2011-03-15 Robert Hogan <robert@webkit.org>
Reviewed by Adam Barth.
Document the qmakearg argument, and hint how to build webkit2 on Qt
https://bugs.webkit.org/show_bug.cgi?id=56326
* Scripts/build-webkit: add --qmakearg to build-webkit --help
2011-03-15 Adam Roben <aroben@apple.com>
Make Windows 7 Release (Tests) a core builder
We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
give it a few more days of greenness first.
Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
Reviewed by Sam Weinig.
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
(BuildBot): Added a regexp to match Windows 7 Release (Tests).
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
(BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
Updated the core regexps to match the buildbot module. Updated the expected results to match
the current set of core builders.
2011-03-14 Adam Roben <aroben@apple.com>
Add forgotten file from r81116
* TestWebKitAPI/Configurations/TestWebKitAPICFNetwork.vsprops: Added.
2011-03-14 Sam Weinig <sam@webkit.org>
Reviewed by Adam Roben.
about:blank fake responses don't get serialized when sent the UIProcess
<rdar://problem/9108119>
https://bugs.webkit.org/show_bug.cgi?id=56357
* TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp: Added.
(TestWebKitAPI::decidePolicyForResponse):
(TestWebKitAPI::TEST):
Add test that loads of about:blank have a response with a MIMEType of text/html.
* TestWebKitAPI/PlatformUtilities.h:
* TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
(TestWebKitAPI::Util::MIMETypeForWKURLResponse):
* TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
(TestWebKitAPI::Util::MIMETypeForWKURLResponse):
Add helper to get the MIMEType from a WKURLResponse.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
Add new files.
2011-03-14 Jarkko Sakkinen <jarkko.j.sakkinen@gmail.com>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Command-line options for QGLWidget and WebGL to QtTestBrowser
https://bugs.webkit.org/show_bug.cgi?id=56006
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::initializeView):
* QtTestBrowser/main.cpp:
(LauncherApplication::handleUserOptions):
2011-03-14 James Kozianski <koz@chromium.org>
Reviewed by Ojan Vafai.
Add method to make BuildBot return test outputs
https://bugs.webkit.org/show_bug.cgi?id=54374
The Builder.latest_cached_build will be used by the new rebaseline
script to determine the build to use baselines from.
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2011-03-14 Adam Roben <aroben@apple.com>
Test that WKBundlePageCanHandleRequest returns true for empty document URLs
Reviewed by Sam Weinig.
* TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp: Added.
(TestWebKitAPI::didReceiveMessageFromInjectedBundle): Store the result of the test.
(TestWebKitAPI::setInjectedBundleClient): Hook up our callback.
(TestWebKitAPI::TEST): Register "emptyscheme" as an empty document scheme, load a page to
ensure the web process is initialized, then ask the bundle to run the test and assert that
it succeeded.
* TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp: Added.
(TestWebKitAPI::CanHandleRequestTest::CanHandleRequestTest): Just call up to the base class.
(TestWebKitAPI::canHandleURL): Helper function to test whether WebKit2 claims to be able to
handle a given URL.
(TestWebKitAPI::runTest): Check that empty document URLs can be handled (and that an unknown
URL type cannot be handled).
(TestWebKitAPI::CanHandleRequestTest::didReceiveMessage): Run the test and return the
result.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
* TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
Added new files to the project.
2011-03-14 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r81094.
http://trac.webkit.org/changeset/81094
https://bugs.webkit.org/show_bug.cgi?id=56355
Broke the chromium DRT related build. (Requested by dave_levin
on #webkit).
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::layerTreeAsText):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::runFileTest):
(TestShell::dump):
* DumpRenderTree/chromium/TestShell.h:
(TestParams::TestParams):
2011-03-14 Daniel Sievers <sievers@google.com>
Reviewed by Simon Fraser.
[Chromium] Make RenderAsTextBehavior and LayerTreeAsTextBehavior tweakable from the DumpRenderTree commandline
https://bugs.webkit.org/show_bug.cgi?id=56139
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(main):
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::layerTreeAsText):
* DumpRenderTree/chromium/LayoutTestController.h:
(LayoutTestController::setShowDebugLayerTree):
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::runFileTest):
(TestShell::dump):
* DumpRenderTree/chromium/TestShell.h:
(TestParams::TestParams):
2011-03-14 Balazs Kelemen <kbalazs@webkit.org>
Reviewed by Adam Roben.
[Qt][WK2]Unbreak InjectedBundle on Qt
https://bugs.webkit.org/show_bug.cgi?id=54109
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::willSendRequestForFrame): Use new WebKit2
API's instead of KURL to analyze the URL.
* WebKitTestRunner/StringFunctions.h:
(WTR::adoptWK): Moved from InjectedBundlePage.cpp to make it available globally
in the WebKitTestRunner project.
(WTR::toWK): Use adoptWK.
(WTR::operator<<): Added helper for printing a WKURL.
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Do not link
against WebCore.
2011-03-14 Balazs Kelemen <kbalazs@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
check-webkit-style falsely should not complain about WebKit2 API headers in the Shared directory
https://bugs.webkit.org/show_bug.cgi?id=56321
* Scripts/webkitpy/style/checker.py: Add the patch "Source/WebKit2/Shared/API/c"
to the list of exceptions for WebKit2 C API headers.
2011-03-14 Adam Roben <aroben@apple.com>
Show the list of recent leaky builds in reverse chronological order
Previously they were being shown in a deterministic but surprising order based on
lexicographically sorting the build offsets (-1, -10, -2, -3...).
Fixes <http://webkit.org/b/56342> Leaks Viewer shows recent builds in a strange order
Reviewed by Darin Adler.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
(RecentBuildsLoader.prototype.start): Sort the builds by revision before passing them to our
callback.
2011-03-14 Adam Roben <aroben@apple.com>
Show a list of up to 10 recent leaky builds when Leaks Viewer loads
Fixes <http://webkit.org/b/56043> Leaks Viewer should present a list of recent builds to
analyze
Reviewed by David Kilzer and Joe Pecoraro.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
(#url-prompt): Made this a bit taller to make room for the list of builds.
(#recent-builds-loading-indicator):
(#recent-builds-list):
Added styles for the build list and loading indicator.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
(LeaksViewer.loaded): Pass the URL from the query through decodeURIComponent before loading
it. This will unescape it.
(LeaksViewer.urlPromptButtonClicked): Moved code from here to _urlChosenFromPrompt.
(LeaksViewer._didLoadRecentBuilds): Added. Stores the builds and updates the prompt.
(LeaksViewer._displayURLPrompt): Start loading the recent leaky builds.
(LeaksViewer._updateURLPrompt): Create a list of the recent leaky builds and add them to the
prompt UI.
(LeaksViewer._urlChosenFromPrompt): Added. Code came from urlPromptButtonClicked. This is
now also used when a recent build is clicked.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js: Added.
(RecentBuildsLoader): Store the callback.
(RecentBuildsLoader.prototype.start): Load the N most recent builds from the specified
builder. Find the revision number, leak count, and results URL for each one. Pass the
resulting info to our callback.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
(range): Added this helper function to return an array containing the integers [0, n).
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added
RecentBuildsLoader.js and added text to the prompt UI to show the recent builds.
2011-03-14 Mark Rowe <mrowe@apple.com>
Reviewed by Adam Roben.
Fix two memory leaks within LayoutTestController's storage support.
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::originsWithLocalStorage): Don't leak the JSStringRef that is returned
by JSStringCreateWithCFString.
(LayoutTestController::deleteLocalStorageForOrigin): Don't leak the WebSecurityOrigin that we
allocate.
2011-03-14 Mark Rowe <mrowe@apple.com>
Reviewed by Oliver Hunt.
<http://webkit.org/b/56304> REGRESSION(r80892): 100,000+ leaks seen on the build bot
* DumpRenderTree/LayoutTestController.cpp:
(LayoutTestController::staticFunctions): Remove duplicate entries.
2011-03-14 Alice Liu <alice.liu@apple.com>
Reviewed by Dan Bernstein.
DumpRenderTree Perl Support fails to build with ARCHS="i386 x86_64"
* DumpRenderTree/mac/PerlSupport/Makefile:
Add whitespace when joining "-arch" with the architectures passed in as arguments.
2011-03-14 Daniel Bates <dbates@rim.com>
https://bugs.webkit.org/show_bug.cgi?id=55438
Add double quotes around value of -DCMAKE_WINCE_SDK that I inadvertently did not add
when landing changeset 81041 <http://trac.webkit.org/changeset/81041>.
* Scripts/build-webkit:
2011-03-14 Daniel Bates <dbates@rim.com>
Reviewed by David Kilzer.
Cleanup: Separate port-specific implementation details from webkitdirs::buildCMakeProject()
https://bugs.webkit.org/show_bug.cgi?id=55438
Separate out the EFL- and WinCE-specific logic from the port-independent logic in
webkitdirs::buildCMakeProject(). Also, remove redundant code for generating the
CMake arguments for feature defines.
* Scripts/build-webkit:
- Added cMakeArgsFromFeatures() to convert the @features array to CMake arguments.
- Modified EFL and WinCE-specific building logic to call buildCMakeProjectOrExit().
- Sorted forward declarations.
- Initialize $makeArgs with the empty string so as to simplify its use in string operations;
Modified call sites as needed.
* Scripts/webkitdirs.pm:
- Added cleanCMakeGeneratedProject() to clean a CMake build.
- Added buildCMakeGeneratedProject() to build using generated build system.
- Added buildCMakeProjectOrExit() which is a facade for building a CMake project.
- Added generateBuildSystemFromCMakeProject() to generate the build system from
a CMake project.
- Removed buildCMakeProject(). This functionality is in buildCMakeProjectOrExit().
2011-03-14 John Knottenbelt <jknotten@chromium.org>
Reviewed by Steve Block.
Detach Geolocation from Frame when Page destroyed.
https://bugs.webkit.org/show_bug.cgi?id=52877
Extend the layout test controller to expose the number of pending
geolocation requests, so that we can test that the requests have
been cancelled on page close.
* DumpRenderTree/LayoutTestController.cpp:
(numberOfPendingGeolocationPermissionRequestsCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/mac/UIDelegate.h:
* DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
2011-03-14 Gyuyoung Kim <gyuyoung.kim@samsung.com>
Unreviewed. Modify my email address in committers.py.
* Scripts/webkitpy/common/config/committers.py:
2011-03-13 Joe Wild <joseph.wild@nokia.com>
Reviewed by Kenneth Rohde Christiansen.
[Qt] QtLauncher does not load the same set of fonts as the DRT
https://bugs.webkit.org/show_bug.cgi?id=34959
This patch adds the option "-use-test-fonts" to the QtTestBrowser.
When this option is used the webkit fonts are loaded the same
as they are in DumpRenderTree. This option can be used on
QtTestBrowser and run-launcher. It can only be used
on Linux systems with FcInit and is configured as such.
* QtTestBrowser/launcherwindow.h:
(WindowOptions::WindowOptions):
* QtTestBrowser/main.cpp:
(initWebKitTestFonts):
(launcherMain):
(LauncherApplication::handleUserOptions):
2011-03-12 Dan Bernstein <mitz@apple.com>
Reviewed by Mark Rowe.
<rdar://problem/8995822> Build DumpRenderTree perl support only for the architectures
DumpRenderTree is being built for.
* DumpRenderTree/mac/PerlSupport/Makefile:
2011-03-12 Dan Bernstein <mitz@apple.com>
Suggested by Mark Rowe.
* DumpRenderTree/mac/Configurations/Base.xcconfig: Leave C++ exceptions enabled
as they were before r80915.
2011-03-12 Anders Carlsson <andersca@apple.com>
Reviewed by Sam Weinig.
Should be able to download a PDF to the Downloads folder
https://bugs.webkit.org/show_bug.cgi?id=56256
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage):
(WTR::TestController::initialize):
Update for API changes.
2011-03-11 Anton D'Auria <adauria@apple.com>
Reviewed and landed by Brady Eidson.
https://bugs.webkit.org/show_bug.cgi?id=51878
Add WebKit1 API to view and delete local storage
Added tests that write to LocalStorage, delete one origin, get list of origins with local storage, delete all origins.
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/LayoutTestController.cpp:
(syncLocalStorageCallback):
(observeStorageTrackerNotificationsCallback):
(deleteAllLocalStorageCallback):
(deleteLocalStorageForOriginCallback):
(originsWithLocalStorageCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/StorageTrackerDelegate.h: Added.
* DumpRenderTree/StorageTrackerDelegate.mm: Added.
(-[StorageTrackerDelegate init]):
(-[StorageTrackerDelegate logNotifications:controller:]):
(-[StorageTrackerDelegate originModified:]):
(-[StorageTrackerDelegate dealloc]):
(-[StorageTrackerDelegate setControllerToNotifyDone:]):
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::deleteAllLocalStorage):
(LayoutTestController::originsWithLocalStorage):
(LayoutTestController::deleteLocalStorageForOrigin):
(observeStorageTrackerNotifications):
(syncLocalStorage):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::originsWithLocalStorage):
(LayoutTestController::deleteAllLocalStorage):
(LayoutTestController::deleteLocalStorageForOrigin):
(LayoutTestController::observeStorageTrackerNotifications):
(LayoutTestController::syncLocalStorage):
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetDefaultsToConsistentValues):
(allocateGlobalControllers):
(releaseGlobalControllers):
* DumpRenderTree/mac/DumpRenderTreeMac.h:
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::syncLocalStorage):
(LayoutTestController::observeStorageTrackerNotifications):
(LayoutTestController::deleteAllLocalStorage):
(LayoutTestController::originsWithLocalStorage):
(LayoutTestController::deleteLocalStorageForOrigin):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::originsWithLocalStorage):
(LayoutTestController::deleteAllLocalStorage):
(LayoutTestController::deleteLocalStorageForOrigin):
(LayoutTestController::observeStorageTrackerNotifications):
(LayoutTestController::syncLocalStorage):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::clearAllApplicationCaches):
(LayoutTestController::syncLocalStorage):
(LayoutTestController::observeStorageTrackerNotifications):
(LayoutTestController::clearAllDatabases):
(LayoutTestController::deleteAllLocalStorage):
(LayoutTestController::originsWithLocalStorage):
(LayoutTestController::deleteLocalStorageForOrigin):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::syncLocalStorage):
(LayoutTestController::observeStorageTrackerNotifications):
(LayoutTestController::clearAllDatabases):
(LayoutTestController::deleteAllLocalStorage):
(LayoutTestController::originsWithLocalStorage):
(LayoutTestController::deleteLocalStorageForOrigin):
2011-03-11 Jessie Berlin <jberlin@apple.com>
Reviewed by Adam Roben.
Crash calling WebContext::clearResourceCaches(InMemoryResourceCachesOnly) before the Web
Process has finished launching.
https://bugs.webkit.org/show_bug.cgi?id=56208
* TestWebKitAPI/Tests/WebKit2/SendingMessagesToTheWebProcessBeforeItIsValid.cpp: Added.
(TestWebKitAPI::TEST):
Test that calling WKContextClearResourceCaches works with both options when called before
the Web Process is valid.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
Add the new test.
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
Ditto.
2011-03-11 Adam Roben <aroben@apple.com>
Make it possible to view all leaks from a build at once in Leaks Viewer
You can now give Leaks Viewer the URL of a build results page (e.g.,
<http://build.webkit.org/results/SnowLeopard%20Intel%20Leaks/r80847%20(15493)/>), and it
will load all the leaks files from that build.
Fixes <http://webkit.org/b/56030> Leaks Viewer: Would like to be able to look at all leaks
files from a particular build at once, rather than one at a time
Reviewed by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksLoader.js: Added.
(LeaksLoader): Initialize our properties.
(LeaksLoader.prototype.start): If the URL ends in .txt, assume it's a single leaks file and
load it. Otherwise assume it's a build results page URL and start fetching the leaks files
from that page.
(LeaksLoader.prototype._loadLeaksFiles): Tell our client how many leaks files we're loading
so it can provide feedback. Then load each one and pass it to the client.
(LeaksLoader.prototype._loadLeaksFromResultsPage): Load the HTML of the results page, parse
it into a DOM, pull out all the links to leaks files, and load each one.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParser.js: Added.
(LeaksParser): Initialize our worker.
(LeaksParser.prototype.addLeaksFile): Tell the worker about the leaks file's contents.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js: Added.
This code mostly came from Worker.js.
(LeaksParserWorker): Initialize our profile.
(LeaksParserWorker.prototype.addLeaksFile): Parse the leaks file and incorporate it into our
profile.
(LeaksParserWorker.prototype._parseLeaks):
(LeaksParserWorker.prototype._createNode):
(LeaksParserWorker.prototype._incorporateLeaks):
Moved these functions here from LeaksViewer. _incorporateLeaks is essentially just a renamed
version of createProfile which adds to this.profile instead of making a new profile each
time.
(onmessage): Parse the file and send back the new profile.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
(#loading-indicator):
(#spinner):
(#loading-indicator-label):
Tweaked styles to accomodate a longer label.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
(LeaksViewer.loaded): Set up a loader and parser.
(LeaksViewer.get filesLeftToParse):
(LeaksViewer.set filesLeftToParse):
Added these simple accessors.
(LeaksViewer._didCountLeaksFiles): Callback from LeaksLoader. Stores the count.
(LeaksViewer._didLoadLeaksFile): Callback from LeaksLoader. Passes the file contents off to
the parser.
(LeaksViewer._didParseLeaksFile): Callback from LeaksParser. If all files have been parsed,
tell the ProfilerAgent and mark that we're done loading. (Code came from the old
_loadLeaksFromURL function).
(LeaksViewer._loadLeaksFromURL): Now just calls through to the loader.
(LeaksViewer._loadingIndicatorText): Added. Returns the text that should show up in the
loading indicator, including the number of files being loaded.
(LeaksViewer._loadingStatusChanged): Update the loading indicator's label, too.
(LeaksViewer._updateLoadingIndicatorLabel): Added. Just updates the label!
(LeaksViewer._updateTitle): Moved code to compute the "Loading" text to
_loadingIndicatorText.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js: Added.
(getResource): Moved here from LeaksViewer.js.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Worker.js: Subsumed by
LeaksParserWorker.js.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added new JS
files, added a #loading-indicator-label element, and tweaked the prompt wording.
2011-03-11 Adam Roben <aroben@apple.com>
Don't trigger a build when build.webkit.org's HTML files are modified
Fixes <http://webkit.org/b/56190> Windows bots shouldn't have built r80848/r80851, but did
Reviewed by Dan Bernstein.
* Scripts/webkitpy/common/config/build.py:
(_should_file_trigger_build): Added build.webkit.org's public_html directory to the list of
directories that shouldn't trigger builds. Re-sorted the list.
* Scripts/webkitpy/common/config/build_unittest.py:
(ShouldBuildTest): Added a test to cover the above change.
2011-03-11 Adam Roben <aroben@apple.com>
Roll out r80848
It was accidentally committed.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
(LeaksViewer._loadLeaksFromURL):
2011-03-11 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
kill hung lighttpd processes on the chromium win bot
https://bugs.webkit.org/show_bug.cgi?id=56152
Sometimes this process doesn't exit and it holds on to
a log file that we can't delete or copy, causing the tests
to go red.
* BuildSlaveSupport/win/kill-old-processes:
2011-03-11 Ilya Sherman <isherman@chromium.org>
Reviewed by Kent Tamura.
[Chromium] missing HTMLInputElement::setValueForUser() to tests
https://bugs.webkit.org/show_bug.cgi?id=55854
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::setValueForUser): Added.
* DumpRenderTree/chromium/LayoutTestController.h:
2011-03-10 Peter Kasting <pkasting@google.com>
Reviewed by Dimitri Glazkov.
Clean up some gross code in TestShellWin.cpp. No functional change.
https://bugs.webkit.org/show_bug.cgi?id=56048
* DumpRenderTree/chromium/TestShellWin.cpp:
(checkLayoutTestSystemDependencies):
2011-03-10 Keith Kyzivat <keith.kyzivat@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] Buildfix for platforms where X11 is not available (e.g. Symbian)
* WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
(WTR::activateFonts): Guard X11-specific code with Q_WS_X11.
2011-03-10 Gyuyoung Kim <gyuyoung.kim@samsung.com>
Reviewed by Kenneth Rohde Christiansen.
[EFL] Make single backing store the default for EWebLauncher
https://bugs.webkit.org/show_bug.cgi?id=55753
Set single backing store the default for EWebLauncher.
* EWebLauncher/main.c:
(main):
2011-03-10 Adam Roben <aroben@apple.com>
Ignore errors from files being held open when creating a test result archive
Fixes <http://webkit.org/b/55581> Test results don't get uploaded if test-result-archive
fails to delete the results directory
Reviewed by Mark Rowe.
* BuildSlaveSupport/test-result-archive:
(archiveTestResults): Ignore errno 2 when deleting the results directory, since it can be
thrown when a process is holding a file open.
2011-03-10 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r80701.
http://trac.webkit.org/changeset/80701
https://bugs.webkit.org/show_bug.cgi?id=56126
fast/dom/Geolocation/window-close-crash.html fails on Snow
Leopard release builds (Requested by mihaip on #webkit).
* DumpRenderTree/LayoutTestController.cpp:
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
* DumpRenderTree/mac/UIDelegate.h:
* DumpRenderTree/mac/UIDelegate.mm:
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2011-03-10 Philippe Normand <pnormand@igalia.com>
Reviewed by Gustavo Noronha Silva.
[GTK] run-gtk-tests doesn't fail if a test crashes
https://bugs.webkit.org/show_bug.cgi?id=56089
* Scripts/run-gtk-tests: gtester -k exits with a 0 exit status
even if a test crashes. So run the tests individually and if a
test fails remember its exit code so run-gtk-test fails too, if
required.
2011-01-26 John Knottenbelt <jknotten@chromium.org>
Reviewed by Dmitry Titov.
Detach Geolocation from Frame when Page destroyed.
https://bugs.webkit.org/show_bug.cgi?id=52877
Extend the layout test controller to expose the number of pending
geolocation requests, so that we can test that the requests have
been cancelled on page close.
* DumpRenderTree/LayoutTestController.cpp:
(numberOfPendingGeolocationPermissionRequestsCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/mac/UIDelegate.h:
* DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
2011-03-09 Adam Roben <aroben@apple.com>
Hide Leaks Viewer's URL prompt by default
We will show it as needed. I think I left it visible by default by accident after doing some
testing.
Fixes <http://webkit.org/b/56031> Leaks Viewer: URL prompt overlay shouldn't appear when a
URL is passed as a query parameter, but does
Reviewed by Simon Fraser.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html:
2011-03-09 Adam Roben <aroben@apple.com>
Don't use Element.classList or box-shadow in Leaks Viewer
They aren't supported by Safari 5. Luckily we have substitutes for them!
Fixes <http://webkit.org/b/56035> Leaks viewer doesn't work in Safari 5
Reviewed by Simon Fraser.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
(#url-prompt): Use -webkit-box-shadow instead of box-shadow.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
(LeaksViewer.urlPromptButtonClicked):
(LeaksViewer._displayURLPrompt):
(LeaksViewer._setLoadingIndicatorHidden):
Use add/removeStyleClass (from utilities.js) instead of Element.classList.
2011-03-09 Adam Roben <aroben@apple.com>
Add a link to the Leaks Viewer tool on the build.webkit.org homepage
Rubber-stamped by David Kilzer.
* BuildSlaveSupport/build.webkit.org-config/templates/root.html:
2011-03-09 Adam Roben <aroben@apple.com>
Add a Leaks Viewer tool to build.webkit.org
This tool parses leaks from the files generated by the leaks bot and displays them in a
view similar to the Web Inspector's Profiles pane (in fact, it uses the Web Inspector's
code).
The URL of the leaks file can be entered in a prompt when the tools is first shown, or can
be passed to the tool via the "url" query parameter.
Only one leaks file can be viewed at a time for now. Only the number of leaks, as opposed to
the number of leaked bytes, is shown. And this tool has made apparent some bugs in the Web
Inspector's Heavy profile view. But it's still pretty useful as-is.
Fixes <http://webkit.org/b/56028> Would like a Shark-like tool for viewing leaks from leak bots
Reviewed by Darin Adler.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css: Added.
(#url-prompt-container): This is the dark overlay that is shown behind the URL prompt.
(#url-prompt): This is the prompt itself.
(#loading-indicator):
(#loading-indicator > img):
These show a spinner in the bottom-right when we're loading a leaks file.
(.percent-time-status-bar-item): Hide the Percent button since we never want to show leaks
as percentages.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js: Added.
(LeaksViewer.loaded): Initialize ourselves and start loading the leaks file (or prompt for a
URL if one wasn't given to us as a query parameter).
(LeaksViewer.get loading):
(LeaksViewer.set loading):
(LeaksViewer.get url):
(LeaksViewer.set url):
Pretty simple accessors.
(LeaksViewer.urlPromptButtonClicked): Start loading the URL from the prompt and hide the
prompt.
(LeaksViewer._displayURLPrompt): Show the prompt and focus the URL input.
(LeaksViewer._loadLeaksFromURL): Start fetching the leaks file. When we've retrieved the
file, hand it off to a Web Worker that will parse it into a fake "profile", then hand off
the profile to the Web Inspector code once it's been created.
(LeaksViewer._loadingStatusChanged): Update our loading indicator and title.
(LeaksViewer._setLoadingIndicatorHidden): Simply hide or show the indicator.
(LeaksViewer._updateTitle): Set our title based on our URL and loading status.
(getResource): Helper function to fetch a resource via XHR and pass it to a callback.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js: Added.
(WebInspector.UIString): Just a simple implementation since this tool isn't localized.
(ProfilerAgent.getProfile): Store the callback.
(ProfilerAgent.profileReady): Call the callback.
(monkeyPatchInspectorObjects): Fix up WebInspector.ProfileDataGridNode to show the self leak
count and total leak counts as real values, not percentages or milliseconds.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Worker.js: Added.
(onmessage): Create a profile based on the leaks file and pass it back to our caller.
(parseLeaks): Extract call stacks out of a leaks file.
(createNode): Helper function to create a new empty profile node.
(createProfile): Take the leak stacks and turn them into a fake "profile" for consumption by
WebInspector.ProfileView.
* BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added.
2011-03-09 Dan Bernstein <mitz@apple.com>
Reviewed by Adam Roben.
Combine multiple --ignore-tests values
https://bugs.webkit.org/show_bug.cgi?id=56016
* Scripts/old-run-webkit-tests:
2011-03-08 Kenji Imasaki <imasaki@chromium.org>
Reviewed by David Levin.
Adding myself to the committer file.
* Scripts/webkitpy/common/config/committers.py:
2011-03-08 Jessie Berlin <jberlin@apple.com>
Rubber-stamped by Adam Roben.
Fix two leaks I introduced into the API tests.
* TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp:
(TestWebKitAPI::TEST):
2011-03-08 Jessie Berlin <jberlin@apple.com>
Reviewed by Anders Carlsson.
Crash in CFNetwork visiting google.com
https://bugs.webkit.org/show_bug.cgi?id=55958
Add a test for not trying to copy a null CFURLResponse.
* TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp: Added.
(TestWebKitAPI::TEST):
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
2011-03-08 Hayato Ito <hayato@chromium.org>
Reviewed by Ojan Vafai.
[NRWT] Renamed a FailureWithType class to a ComparisonTestFailure class.
FailureWithType is a confusing naming because we have removed test_type/* classes in r79840.
https://bugs.webkit.org/show_bug.cgi?id=55450
* Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
2011-03-08 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] Add layoutTestController.setValueForUser() after r80412
https://bugs.webkit.org/show_bug.cgi?id=55862
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::setValueForUser): Implement setValueForUser()
2011-03-08 Patrick Gansterer <paroga@webkit.org>
Reviewed by Adam Barth.
Add script_shell_command to WebKitPort
https://bugs.webkit.org/show_bug.cgi?id=55925
Add a central place where we generate the shell command from the script name.
This function will be used later to detect the correct script interpreter on Win32.
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/tool/steps/abstractstep.py:
* Scripts/webkitpy/tool/steps/preparechangelog.py:
2011-03-07 Ojan Vafai <ojan@chromium.org>
Reviewed by Adam Barth.
add webkit-patch attach-to-bug for posting to bugzilla from the commandline
https://bugs.webkit.org/show_bug.cgi?id=55749
* Scripts/webkitpy/tool/commands/upload.py:
* Scripts/webkitpy/tool/commands/upload_unittest.py:
* Scripts/webkitpy/tool/steps/__init__.py:
* Scripts/webkitpy/tool/steps/attachtobug.py: Added.
* Scripts/webkitpy/tool/steps/options.py:
2011-03-07 Daniel Cheng <dcheng@chromium.org>
Reviewed by David Levin.
Add support for DataTransferItems
https://bugs.webkit.org/show_bug.cgi?id=55115
Enable data transfer items in DRT.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):
2011-03-06 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
[test dashboard] pull treemap files
https://bugs.webkit.org/show_bug.cgi?id=55852
* TestResultServer/handlers/dashboardhandler.py:
2011-03-03 Hayato Ito <hayato@chromium.org>
Reviewed by Mihai Parparita.
[NRWT] Clear output image file used by DumpRenderTree to make sure the
previous image should not be used in the ChromiumDriver.
Also make sure that output image from the ChromiumDriver should be ''
(empty string) if a test crashes for consistency with the WebKitDriver
implementation.
https://bugs.webkit.org/show_bug.cgi?id=55746
* Scripts/webkitpy/layout_tests/port/chromium.py:
2011-03-07 Zan Dobersek <zandobersek@gmail.com>
Reviewed by Martin Robinson.
[Gtk] fast/js/navigator-language.html fails locally
https://bugs.webkit.org/show_bug.cgi?id=55880
Add LANG variable to the clean environment when running
old-run-webkit-tests script.
* Scripts/old-run-webkit-tests:
2011-03-07 Tony Chang <tony@chromium.org>
Reviewed by Mihai Parparita.
[Chromium] Find .checksum files without .pngs and vice-versa
https://bugs.webkit.org/show_bug.cgi?id=55236
* Scripts/find-mismatched-layout-test-results: Added. A script for finding
mismatched results.
2011-03-07 Chris Fleizach <cfleizach@apple.com>
Reviewed by Beth Dakin.
AX: WK1 needs to use the ScrollView attachment for AXScrollArea, WK2 should not use this element.
https://bugs.webkit.org/show_bug.cgi?id=55706
The rootElement in DRT should point to the scroll view. In WK1, that will be the parent of the rootObject.
* DumpRenderTree/mac/AccessibilityControllerMac.mm:
(AccessibilityController::focusedElement):
Remove FIXME comment about caching the focusedElement. In fact, we shouldn't cache this
because it could change at any time.
(AccessibilityController::rootElement):
2011-03-07 Sam Weinig <sam@webkit.org>
Reviewed by Anders Carlsson.
Add unableToImplementPolicy callback for WebKit2 policy client
<rdar://problem/9071902>
https://bugs.webkit.org/show_bug.cgi?id=55884
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]):
Update policy client initialization struct for new member.
2011-03-07 Sam Weinig <sam@webkit.org>
Reviewed by Anders Carlsson.
Replace WebKit2's decidePolicyForMIMEType with decidePolicyForResponse
https://bugs.webkit.org/show_bug.cgi?id=55827
* MiniBrowser/mac/BrowserWindowController.m:
(decidePolicyForResponse):
(-[BrowserWindowController awakeFromNib]):
* TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
(TestWebKitAPI::decidePolicyForResponse):
(TestWebKitAPI::TEST):
Update tools for new name.
2011-03-06 Daniel Bates <dbates@rim.com>
Reviewed by Eric Seidel and David Kilzer.
Fix misspelled word in build-webkit's checkForJavaSDK() and sort forward declarations
https://bugs.webkit.org/show_bug.cgi?id=55503
Fix misspelled word "Dowloads" [sic] in the message printed to standard output
in checkForJavaSDK().
Also, remove extraneous space in function prototype for unlinkZeroFiles.
* Scripts/build-webkit:
2011-03-05 Brian Weinstein <bweinstein@apple.com>
Reviewed by Sam Weinig.
Fix Web Process crashes caused by https://bugs.webkit.org/show_bug.cgi?id=53919.
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::willSendRequestForFrame): Retain the WKURLRequestRef returned
by willSendRequestForFrame. The API expects a retained reference to the URL request.
2011-03-05 Ilya Sherman <isherman@chromium.org>
Reviewed by Darin Adler.
HTMLInputElement::setValue() should schedule change event when the element is focused
In service of https://code.google.com/p/chromium/issues/detail?id=42716
https://bugs.webkit.org/show_bug.cgi?id=53160
* DumpRenderTree/LayoutTestController.cpp:
(setValueForUserCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::setValueForUser): Stubbed out; need real implementation.
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::setValueForUser): Expose HTMLInputElement::setValueForUser() to tests
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setValueForUser): Stubbed out; need real implementation.
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::setValueForUser): Stubbed out; need real implementation.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::setValueForUser): Stubbed out; need real implementation.
2011-03-04 Mark Rowe <mrowe@apple.com>
Reviewed by Alexey Proskuryakov.
<rdar://problem/9055252> DumpRenderTree is crashing with GuardMalloc
GuardMalloc does not expect the memory containing its malloc_zone_t to ever be read-only.
* DumpRenderTree/mac/CheckedMalloc.cpp:
(protectionOfRegion): Retrieve the protection flags for the VM region containing the given address.
(makeLargeMallocFailSilently): Save the initial protection flags, temporarily make the region writeable,
and then restore the initial protection flags when we're done.
2011-03-04 Evan Martin <evan@chromium.org>
Reviewed by Tony Chang.
[gdb] pretty-print KURL in gdb
https://bugs.webkit.org/show_bug.cgi?id=55788
* gdb/webkit.py: extend the pretty-printers to print WTF::CString
and KURLGooglePrivate. As always, "p/r" can be used to sidestep
pretty-printing.
2011-03-03 Timothy Hatcher <timothy@apple.com>
Update the tests that use WKPageRunJavaScriptInMainFrame use the WKSerializedScriptValueRef.
Reviewed by Darin Adler.
* TestWebKitAPI/JavaScriptTest.cpp:
(TestWebKitAPI::javaScriptCallback):
* TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp:
(TestWebKitAPI::didRunJavaScript):
* TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
(TestWebKitAPI::nullJavaScriptCallback):
* TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp:
(TestWebKitAPI::didRunJavaScript):
2011-03-03 Timothy Hatcher <timothy@apple.com>
Make run-api-tests show failed if a test fails.
https://webkit.org/b/55725
Reviewed by Sam Weinig.
* Scripts/run-api-tests:
(runTest): Remove "my" from the nested $result assignment so the
check outside the condition sees the real result.
2011-03-02 MORITA Hajime <morrita@google.com>
Reviewed by Kent Tamura.
[Chromium][DRT] WebTask should use webkit_support::TaskAdaptor
https://bugs.webkit.org/show_bug.cgi?id=55434
Changed WebTask to be a subclass of webkit_support::TaskAdaptor
and passed the object webkit_support::PostDelayedTask directly.
Note that the change keeps postTask() as is because it has a
semantics taht webkit_support::PostDelayedTask() has; It jumps to
the main thread.
* DumpRenderTree/chromium/Task.cpp:
(postDelayedTask):
* DumpRenderTree/chromium/Task.h:
(WebTask::Run):
2011-03-03 Mahesh Kulkarni <mahesh.kulkarni@nokia.com>
Reviewed by Kenneth Rohde Christiansen.
[QT] Implement mock client-based geolocation for layout testing
https://bugs.webkit.org/show_bug.cgi?id=54334
Implement testing for Client-Based geolocation for QtWebkit.
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::DumpRenderTree):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::reset):
(LayoutTestController::setGeolocationPermission):
(LayoutTestController::setMockGeolocationError):
(LayoutTestController::setMockGeolocationPosition):
2011-03-03 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r80237.
http://trac.webkit.org/changeset/80237
https://bugs.webkit.org/show_bug.cgi?id=55714
Windows bots aren't ready to be core (Requested by aroben on
#webkit).
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2011-03-03 Mihai Parparita <mihaip@chromium.org>
Reviewed by Tony Chang.
NRWT: AttributeError: TestRunner2 instance has no attribute '_cancel_workers'
https://bugs.webkit.org/show_bug.cgi?id=55694
Fixes for interruptions in NRWT:
- Make TestRunInterruptedException be pickleable correctly (the base
Exception class defines a __reduce__ that does not include the reason)
- Fix ordering of returned arguments from TestRunner2._run_tests
(interrupted and keyboard_interrupted were reversed)
- Fix cancel_workers callsites (was using old name).
- In handle_exception re-raise actual exception instance that was thrown
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
* Scripts/webkitpy/layout_tests/layout_package/worker.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-03-03 Qi Zhang <qi.2.zhang@nokia.com>
Unreviewed.
Adding myself as a committer.
* Scripts/webkitpy/common/config/committers.py:
2011-03-03 Tony Gentilcore <tonyg@chromium.org>
Reviewed by Mihai Parparita.
Teach sheriffbot to answer whois commands
https://bugs.webkit.org/show_bug.cgi?id=55687
* Scripts/webkitpy/tool/bot/irc_command.py:
* Scripts/webkitpy/tool/bot/irc_command_unittest.py:
* Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2011-03-03 Victoria Kirst <vrk@google.com>
Reviewed by Eric Carlson.
Enable media tests in chromium-gpu/test_expectations
https://bugs.webkit.org/show_bug.cgi?id=53608
Add media folder to the list of GPU test paths.
* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2011-03-03 Adam Roben <aroben@apple.com>
Work around a bug in KURL's parsing of Windows-style absolute file: URLs
Fixes <http://webkit.org/b/55674> Many tests fail in WebKit2 mode on Windows due to
incorrect parsing of absolute Windows-style file: URLs
Reviewed by Tim Hatcher.
* WebKitTestRunner/TestInvocation.cpp:
(WTR::createWKURL): Add a "localhost" host to file: URLs we create to work around
<http://webkit.org/b/55683>.
2011-03-03 Tony Chang <tony@chromium.org>
Reviewed by Dimitri Glazkov.
[chromium] Run kill-old-processes on the chromium-win bot too
https://bugs.webkit.org/show_bug.cgi?id=55630
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
* BuildSlaveSupport/win/kill-old-processes: Add wdiff.exe to the kill list.
2011-03-03 Carlos Garcia Campos <cgarcia@igalia.com>
Reviewed by Martin Robinson.
[GTK] DRT needs implementation of EventSender.scheduleAsynchronousClick
https://bugs.webkit.org/show_bug.cgi?id=53960
* DumpRenderTree/gtk/EventSender.cpp:
(sendClick):
(scheduleAsynchronousClickCallback):
2011-03-03 Adam Roben <aroben@apple.com>
Look for Windows-style paths when checking whether a test is an Inspector test
WTR part of <http://webkit.org/b/55672> <rdar://problem/9080867> All inspector tests time
out or crash on Windows 7 Release (WebKit2 Tests)
Reviewed by Darin Adler.
* WebKitTestRunner/TestInvocation.cpp:
(WTR::shouldOpenWebInspector): Also look for backslashes, since that's what we'll get on
Windows.
2011-03-03 Adam Roben <aroben@apple.com>
Make Windows 7 Release (Tests) a core builder
We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
give it a few more days of greenness first.
Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
Reviewed by Andreas Kling.
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
(BuildBot): Added a regexp to match Windows 7 Release (Tests).
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
(BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
Updated the core regexps to match the buildbot module. Updated the expected results to match
the current set of core builders.
2011-03-03 Benjamin Poulain <benjamin.poulain@nokia.com>
Reviewed by Andreas Kling.
Adding myself as a reviewer.
* Scripts/webkitpy/common/config/committers.py:
2011-03-03 Lukasz Slachciak <l.slachciak@samsung.com>
Reviewed by Xan Lopez.
[GTK] Extended application cache database API and added unit tests file.
https://bugs.webkit.org/show_bug.cgi?id=55335
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: removed external declaration
of webkit_application_cache_set_maximum_size due to API exposal.
2011-03-02 Ojan Vafai <ojan@chromium.org>
Reviewed by Adam Barth.
make webkit-patch upload work for security bugs
https://bugs.webkit.org/show_bug.cgi?id=55655
The problem is that we shell out to prepare-ChangeLog,
which does not authenticate with bugzilla. All it needs from
bugzilla is the bug title though. We just pass it through from
webkit-patch instead.
* Scripts/prepare-ChangeLog:
* Scripts/webkitpy/common/net/bugzilla/bug.py:
* Scripts/webkitpy/tool/steps/preparechangelog.py:
2011-03-01 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
[test-dashboard] make gtest JSON match layout-tests JSON
https://bugs.webkit.org/show_bug.cgi?id=55556
http://trac.webkit.org/changeset/80090 stopped converting time_ms to t.
Make this code match.
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2011-03-02 Dirk Pranke <dpranke@chromium.org>
Reviewed by Tony Chang.
NRWT - implement Linux Hardy 64-bit port support. This involves
adding support for non-'x86' architectures and updating the
"deduplicate_tests" script to correctly understand that
port.name() is not always equal to basename(port.baseline_path()).
This change introduces two new port names: 'chromium-linux-x86'
and 'chromium-linux-x86_64'. Using the prior 'chromium-linux'
flag will cause the code to determine which one to use at
runtime, just like we currently do with 'chromium-win'.
https://bugs.webkit.org/show_bug.cgi?id=55535
* Scripts/webkitpy/common/system/executive_mock.py:
Handle 'return_stderr' keyword arg to run_command()
* Scripts/webkitpy/layout_tests/deduplicate_tests.py:
* Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
* Scripts/webkitpy/layout_tests/port/factory.py:
* Scripts/webkitpy/layout_tests/port/google_chrome.py:
2011-03-02 Kevin Ollivier <kevino@theolliviers.com>
[wx] Build fix, make sure we put the wxWebKit dylib in the right directory and
update symlinks accordingly.
* wx/packaging/build-mac-installer.py:
2011-03-02 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r80139.
http://trac.webkit.org/changeset/80139
https://bugs.webkit.org/show_bug.cgi?id=55620
failing and crashing tests on Snow Leopard bot (Requested by
mihaip1 on #webkit).
* DumpRenderTree/LayoutTestController.cpp:
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
* DumpRenderTree/mac/UIDelegate.h:
* DumpRenderTree/mac/UIDelegate.mm:
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2011-03-02 Adam Roben <aroben@apple.com>
Wait for force a paint in WebKitTestRunner until we're actually ready to dump the output
This matches DumpRenderTree on Windows.
Fixes <http://webkit.org/b/55469> <rdar://problem/9068539> REGRESSION (r79863): Lots of
dom/html/level2/html/HTMLFrameElement*.html tests crashing in FrameView::paintOverhangAreas
in WebKit2
Reviewed by Beth Dakin.
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::dump): Moved code to force a paint here...
(WTR::InjectedBundlePage::didFinishLoadForFrame): ...from here.
2011-03-02 Daniel Cheng <dcheng@chromium.org>
Reviewed by David Levin.
Add feature define for data transfer items
https://bugs.webkit.org/show_bug.cgi?id=55510
* Scripts/build-webkit:
2011-03-02 Adam Roben <aroben@apple.com>
Look in the new-to-SnowLeopard location for crash logs on SnowLeopard and newer
ReportCrash saves logs to ~/Library/Logs/CrashReporter on Leopard, but to
~/Library/Logs/DiagnosticReports on SnowLeopard. old-run-webkit-tests was only looking in
the former location, but was getting lucky on some SnowLeopard machines because symlinks
were being created there pointing into DiagnosticReports. For machines without these
symlinks, crash logs were not getting captured.
Fixes <http://webkit.org/b/55607> old-run-webkit-tests doesn't capture crash logs on some
SnowLeopard machines
Reviewed by Eric Seidel.
* Scripts/old-run-webkit-tests:
(captureSavedCrashLog): Look in ~/Library/Logs/DiagnosticReports on SnowLeopard and newer.
2011-03-02 Adam Roben <aroben@apple.com>
Don't try to capture crash logs on platforms that haven't implemented that feature
Fixes <http://webkit.org/b/55504> Use of uninitialized value in numeric lt (<) at
Tools/Scripts/old-run-webkit-tests line 1778 seen on Qt bots after a crash
Reviewed by Joseph Pecoraro.
* Scripts/old-run-webkit-tests:
(captureSavedCrashLog): Bail if we don't have a crash log glob.
(findNewestFileMatchingGlob): Fix a buggy test for when no paths match the glob. This isn't
strictly required to fix bug 55504, but seems worthwhile.
2011-01-26 John Knottenbelt <jknotten@chromium.org>
Reviewed by Dmitry Titov.
Detach Geolocation from Frame when Page destroyed.
https://bugs.webkit.org/show_bug.cgi?id=52877
Extend the layout test controller to expose the number of pending
geolocation requests, so that we can test that the requests have
been cancelled on page close.
* DumpRenderTree/LayoutTestController.cpp:
(numberOfPendingGeolocationPermissionRequestsCallback):
(LayoutTestController::staticFunctions):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/chromium/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/chromium/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/mac/UIDelegate.h:
* DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::numberOfPendingGeolocationPermissionRequests):
2011-03-01 Kent Tamura <tkent@chromium.org>
Reviewed by Dimitri Glazkov.
Assertion fails when a form validation bubble appears
https://bugs.webkit.org/show_bug.cgi?id=55550
Test: fast/forms/interactive-validation-attach-assertion.html
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::applyTo): Enable the interactive validation feature by
default. This matches to Chromium browser.
2011-03-01 Gyuyoung Kim <gyuyoung.kim@samsung.com>
Reviewed by Kent Tamura.
Remove efl port style exception of Webcore/JavaScriptCore
https://bugs.webkit.org/show_bug.cgi?id=55549
WebCore and JavaScriptCore need to adhere WebKit coding style.
* Scripts/webkitpy/style/checker.py:
2011-03-01 Tony Chang <tony@chromium.org>
Reviewed by Ojan Vafai.
[chromium] clean up temp files on chromium-mac builder too
https://bugs.webkit.org/show_bug.cgi?id=55525
* BuildSlaveSupport/chromium/remove-crash-logs:
2011-02-28 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
wrap json in a function call to afford cross-domain loading
https://bugs.webkit.org/show_bug.cgi?id=55353
Also delete the code that compacts the JSON. It turns out this
breaks the rebaseline tool and is probably a premature optimization anyways.
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
* Scripts/webkitpy/tool/commands/rebaselineserver.py:
2011-03-01 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r80079.
http://trac.webkit.org/changeset/80079
https://bugs.webkit.org/show_bug.cgi?id=55547
"Broke the Win debug build?" (Requested by dcheng on #webkit).
* Scripts/build-webkit:
2011-03-01 Mark Rowe <mrowe@apple.com>
Reviewed by Sam Weinig.
Update verification scripts to handle WebKit2.
* Scripts/check-for-global-initializers: Skip object files that use RefCountedLeakChecker.
* Scripts/check-for-inappropriate-files-in-framework: Add some inappropriate file types for
WebKit2.
2011-03-01 Daniel Cheng <dcheng@chromium.org>
Reviewed by David Levin.
Add feature define for data transfer items
https://bugs.webkit.org/show_bug.cgi?id=55510
* Scripts/build-webkit:
2011-02-23 Joseph Pecoraro <joepeck@webkit.org>
Reviewed by Timothy Hatcher.
All Console Messages should be passed to ChromeClients.
https://bugs.webkit.org/show_bug.cgi?id=54926
Update DumpRenderTree's UIDelegate to use the new console message delegate.
* DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate webView:addMessageToConsole:withSource:]):
2011-03-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Barth.
Tragically, we stopped running the python unit tests
on Chromium Win just as I was getting them to be green.
This patch turns them back on.
https://bugs.webkit.org/show_bug.cgi?id=55521
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
2011-03-01 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
Disable
webkitpy.common.system.executive_unittest.ExecutiveTest.\
test_kill_all fails on win until it can be properly triaged.
https://bugs.webkit.org/show_bug.cgi?id=54790
* Scripts/webkitpy/common/system/executive_unittest.py:
2011-03-01 Helder Correia <helder@sencha.com>
Unreviewed. Add myself to committers.py.
* Scripts/webkitpy/common/config/committers.py:
2011-03-01 Adam Roben <aroben@apple.com>
Give up if a crash log for the web process is taking too long to be saved on Windows
This should keep the bots from getting stuck while waiting for a crash log, if writing a
crash log ever hangs or the UI process for some reason doesn't notice it's finished.
Fixes <http://webkit.org/b/55499> WebKitTestRunner can hang forever waiting for a crash log
to be saved for the web process
Reviewed by Steve Falkenburg.
* WebKitTestRunner/win/TestControllerWin.cpp:
(WTR::runRunLoopUntil): Moved code to run the run loop here here from platformRunUntil.
Generalized the code slightly to handle the optional object to wait on.
(WTR::TestController::platformRunUntil): Use the new runRunLoopUntil function, and also use
it (with a timeout) when waiting for a crash log for the web process to be saved. This will
prevent us from waiting forever.
2011-03-01 Adam Roben <aroben@apple.com>
Give WebKitLauncherWin's PDB file a name different from WebKit's
WebKitLauncherWin (which builds WebKit.exe) and WebKit (which builds WebKit.dll) were
fighting over the same PDB file, meaning that one of them was going to lose and not have
symbols. WebKitLauncherWin's PDB file is now named WebKitLauncherWin.pdb.
Fixes <http://webkit.org/b/55495> Crashes in WebKit.dll have bad backtraces for some test
runs (due to missing symbols)
Reviewed by David Kilzer.
* WebKitLauncherWin/WebKitLauncherWinCommon.vsprops: Use the project name,
"WebKitLauncherWin", instead of the target name, "WebKit", when naming the PDB file.
2011-03-01 Adam Roben <aroben@apple.com>
Count web process crashes toward the --exit-after-n-crashes-or-timeouts limit
Fixes <http://webkit.org/b/55492> Web process crashes aren't counted toward the
--exit-after-n-crashes-or-timeouts limit
Reviewed by David Kilzer.
* Scripts/old-run-webkit-tests:
(stopRunningTestsEarlyIfNeeded): Include web process crashes in the computation, and print
the number of tests that caused the web process to crash when exiting early.
2011-03-01 Adam Roben <aroben@apple.com>
Save a crash log when the web process crashes
On Windows, WebKitTestRunner now detects when the web process is crashing and waits to exit
until it has finished crashing, which guarantees that the crash log will have had time to be
saved, too. On Mac, we always wait until ReportCrash has exited before capturing the crash
log, so all we have to do is choose the right crash log out of the CrashReporter directory.
Fixes <http://webkit.org/b/44121> <rdar://problem/8320759> When the web process crashes and
a crash log is being saved, WebKitTestRunner thinks the web process has become unresponsive
Reviewed by Sam Weinig.
* Scripts/old-run-webkit-tests:
(testCrashedOrTimedOut): Don't kill WebKitTestRunner when the web process crashes. It will
kill itself. On Windows, this will cause us to wait until the crash log has been saved. On
Mac, it should have no effect. Capture saved crash logs for web process crashes, too.
(captureSavedCrashLog): Added $webProcessCrashed argument. On Mac, look for
WebProces_*.crash files when the web process crashes.
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::initialize): Added an initializationUserData argument. Updated for
initializePlatformDefaults -> platformInitialize rename. Pass the initializationUserData
along to platformInitialize.
* WebKitTestRunner/InjectedBundle/InjectedBundle.h: See above.
* WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
(WKBundleInitialize): Pass along the initializationUserData to the InjectedBundle.
* WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
(WTR::InjectedBundle::platformInitialize):
* WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
(WTR::InjectedBundle::platformInitialize):
Updated function signature.
* WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
(WTR::exceptionFilter): Added. Tells the UI process we're crashing by signaling the
webProcessCrashingEvent, then lets the crash continue as normal.
(WTR::InjectedBundle::platformInitialize): Hook up exceptionFilter. Retrieve the name of the
event we should use to tell the UI process we're crashing from the initializationUserData,
and get a handle to that event.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::TestController): Initialize new members.
(WTR::TestController::processDidCrash): Removed unnecessary WKPageRef argument. Changed to
only print the "#CRASHED - WebProcess" message once, since this can be called more than once
when a crash log is being saved on Windows. Exit right away if specified. (This is the
default.)
* WebKitTestRunner/TestController.h: Added new members.
* WebKitTestRunner/win/TestControllerWin.cpp:
(WTR::TestController::platformInitialize): Set up the event the web process will use to tell
us it's crashing.
(WTR::TestController::platformRunUntil): Pass MWMO_INPUTAVAILABLE to
::MsgWaitForMultipleObjectsEx so we'll process messages that have already been seen by
::PeekMessage. (This is unrelated to the bug fix.) Notice when the webProcessCrashingEvent
has been signaled. When this happens, print the "#CRASHED - WebProcess" message right away
so the test harness will know the web process has crashed and not try to kill us, then wait
for the web process to finish crashing so a crash log will have time to be saved.
(WTR::toWK): Simple hepler function.
(WTR::TestController::platformInitializeContext): Pass along the name of the event the web
process should use to tell us it is crashing in the context's initialization user data.
2011-03-01 Dimitri Glazkov <dglazkov@chromium.org>
Reviewed by Tony Gentilcore.
Move the checks to exclude Python/Perl tests for Chromium Win build to the right place.
https://bugs.webkit.org/show_bug.cgi?id=55476
Chromium Win uses BuildAndTestFactory, not TestFactory.
* BuildSlaveSupport/build.webkit.org-config/master.cfg: Moved to the right place.
2011-03-01 Adam Roben <aroben@apple.com>
Test that the WebKit2 responsiveness timer doesn't fire too early
Test for <http://webkit.org/b/55417> <rdar://problem/9065287> RunLoop::Timer fires 1000x too
early on Windows
Reviewed by Anders Carlsson.
* TestWebKitAPI/PlatformUtilities.h: Added sleep().
* TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp: Added.
(TestWebKitAPI::didReceiveMessageFromInjectedBundle):
(TestWebKitAPI::didFinishLoadForFrame):
(TestWebKitAPI::processDidBecomeUnresponsive):
(TestWebKitAPI::setInjectedBundleClient):
(TestWebKitAPI::setPageLoaderClient):
Simple helper functions.
(TestWebKitAPI::TEST): Load an HTML file to make sure the web process is initialized. Then
tell the web process to pause and press the spacebar key. The spacebar keypress should cause
the responsiveness timer to start, but the web process should unpause before it has a chance
to fire. Run until the web process has unpaused, and assert that the timer didn't fire.
* TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly_Bundle.cpp: Added.
(TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::ResponsivenessTimerDoesntFireEarlyTest):
Call up to the base class.
(TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::didReceiveMessage): When asked to
pause, sleep for 0.5 seconds, then send back a message saying we paused.
* TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
(TestWebKitAPI::Util::sleep): Added. Calls through to usleep.
* TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
(TestWebKitAPI::Util::sleep): Added. Calls through to ::Sleep.
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
* TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
Added new files.
2011-03-01 Andras Becsi <abecsi@webkit.org>
Reviewed by Csaba Osztrogonác.
[Qt] Clean up the project files and move common options to WebKit.pri.
* WebKitTestRunner/qt/WebKitTestRunner.pro: Deduplicate.
2011-03-01 Adam Roben <aroben@apple.com>
Skip another multiprocessing test on Windows
It is sometimes failing (and may be leaving child Python processes around).
<http://webkit.org/b/55087> tracks the failure.
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
(FuncitonTests.test_get__processes): Skip this test on Windows.
2011-03-01 Ilya Tikhonovsky <loislo@chromium.org>
Reviewed by Pavel Feldman.
Web Inspector: reduce inspector tests flakiness.
https://bugs.webkit.org/show_bug.cgi?id=55401
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::runFileTest):
* DumpRenderTree/chromium/WebPreferences.cpp:
(WebPreferences::reset):
2011-02-28 Mihai Parparita <mihaip@chromium.org>
Reviewed by James Robinson.
NRWT should still upload results when it exits early due to too many failures
https://bugs.webkit.org/show_bug.cgi?id=55432
Only keyboard interrupts should prevent results from being uploaded.
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-02-28 Alice Liu <alice.liu@apple.com>
Reviewed by Adam Roben and Gavin Barraclough.
Test for https://bugs.webkit.org/show_bug.cgi?id=54898
InjectedBundleNodeHandle dies too early in WKBundleHitTestResultGetNodeHandle.
* TestWebKitAPI/InjectedBundleController.h:
(TestWebKitAPI::InjectedBundleController::bundle):
* TestWebKitAPI/PlatformWebView.h:
* TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp: Added.
(TestWebKitAPI::didFinishLoadForFrame):
(TestWebKitAPI::didReceiveMessageFromInjectedBundle):
(TestWebKitAPI::setPageLoaderClient):
(TestWebKitAPI::TEST):
* TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp: Added.
(TestWebKitAPI::HitTestResultNodeHandleTest::HitTestResultNodeHandleTest):
(TestWebKitAPI::HitTestResultNodeHandleTest::getContextMenuFromDefaultMenu):
(TestWebKitAPI::HitTestResultNodeHandleTest::didCreatePage):
* TestWebKitAPI/mac/PlatformWebViewMac.mm:
(TestWebKitAPI::PlatformWebView::simulateRightClick):
* TestWebKitAPI/win/PlatformWebViewWin.cpp:
(TestWebKitAPI::PlatformWebView::simulateRightClick):
Added files to:
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
* TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
2011-02-28 Adam Roben <aroben@apple.com>
Start running InjectedBundle API tests on Windows
A new project, TestWebKitAPIInjectedBundle, builds the injected bundle DLL.
Fixes <http://webkit.org/b/55420> run-api-tests doesn't run injected bundle tests on Windows
Reviewed by Darin Adler.
* TestWebKitAPI/Configurations/TestWebKitAPIDebug.vsprops: Added.
* TestWebKitAPI/Configurations/TestWebKitAPIDebugAll.vsprops: Added.
* TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops: Added.
* TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops: Added.
* TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops: Added.
These are just extracted from TestWebKitAPI.vcproj.
* TestWebKitAPI/Configurations/TestWebKitAPIInjectedBundleCommon.vsprops: Added. Just sets
up the DLL's filename.
* TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Fixed DLL name.
(TestWebKitAPI::Util::createInjectedBundlePath): Changed to use kCFAllocatorNull instead of
0 (== kCFAllocatorDefault) so that CF won't try to deallocate our C string.
* TestWebKitAPI/win/TestWebKitAPI.sln: Added TestWebKitAPIInjectedBundle. Removed the bogus
"all" configuration.
* TestWebKitAPI/win/TestWebKitAPI.vcproj: Changed to use the new .vsprops files. Added
injected bundle tests. Let VS resort things.
* TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Copied from Tools/TestWebKitAPI/win/TestWebKitAPI.vcproj.
2011-02-28 Adam Roben <aroben@apple.com>
Add a test for the paint rect passed via WM_PAINT to windowless plugins
Test for <http://webkit.org/b/55365> <rdar://problem/9031089> REGRESSION (r79040): WebKit2:
Large portions of pages with plugins paint black
Reviewed by Anders Carlsson.
* DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
(PluginTest::NPP_HandleEvent):
* DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
Added NPP_HandleEvent. Just returns 0 at this level.
* DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowlessPaintRectCoordinates.cpp: Added.
(WindowlessPaintRectCoordinates::WindowlessPaintRectCoordinates): Call up to the base class.
(WindowlessPaintRectCoordinates::NPP_New): Mark ourselves as windowless.
(WindowlessPaintRectCoordinates::NPP_HandleEvent): Check that the paint rect passed via
WM_PAINT has the expected coordinates.
* DumpRenderTree/TestNetscapePlugIn/main.cpp:
(NPP_HandleEvent): Give the PluginTest a chance to handle the event before doing anything
else.
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new file, let VS
resort things.
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::didFinishLoadForFrame): Added a call to WKBundlePageForceRepaint.
This is roughly equivalent to the call to -[WebView displayIfNeeded] in -[FrameLoadDelegate
webView:didFinishLoadForFrame:] in DumpRenderTree.
* WebKitTestRunner/win/PlatformWebViewWin.cpp:
(WTR::PlatformWebView::PlatformWebView): Tell the WKView it's in a window so that plugins
can start running.
2011-02-28 Adam Roben <aroben@apple.com>
Tell MiniBrowser's WKViews that they're in a window
Fixes <http://webkit.org/b/55364> Plugins don't work in MiniBrowser on Windows
Reviewed by Anders Carlsson.
* MiniBrowser/win/BrowserView.cpp:
(BrowserView::create): Call WKViewSetIsInWindow after creating the view.
2011-02-28 Alexis Menard <alexis.menard@openbossa.org>
Reviewed by Csaba Osztrogonác.
The new QML tests require QML_IMPORT_PATH variable to be set in order to
load the proper plugin.
* Scripts/run-qtwebkit-tests:
2011-02-28 Joone Hur <joone.hur@collabora.co.uk>
Unreviewed, fix spelling mistake.
hardware-accelated => hardware-accelerated.
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-02-28 Kristian Amlie <kristian.amlie@nokia.com>
Reviewed by Andreas Kling.
Made sure that the build-webkit --qmake argument is always respected.
Previously, it would be respected during the build, but not during
the early feature detection. The build would also fail if qmake was
not in your path, even if you specified --qmake with a valid qmake
binary.
[Qt] WebKit patches required to work with a modularized version of Qt
https://bugs.webkit.org/show_bug.cgi?id=53916
* Scripts/build-webkit:
* Scripts/webkitdirs.pm:
2011-02-27 Ojan Vafai <ojan@chromium.org>
Fix failing tests from http://trac.webkit.org/changeset/79837.
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
This was just a mistake that would break if someone tried printing with this configuration.
* Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
Method signature changed. Didn't notice that this method was being calling in a unittest.
2011-02-24 Hayato Ito <hayato@chromium.org>
Reviewed by Eric Seidel.
[NRWT] Remove test_types/* classes.
This is the last patch in the series of efforts which tried to remove test_types classes.
A single_test_runner module will take over responsibilities of test_type classes.
From this patch, we won't display statistics per test_type even if a '--verbose' option is given.
https://bugs.webkit.org/show_bug.cgi?id=55123
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_results.py:
* Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
* Scripts/webkitpy/layout_tests/test_types/__init__.py: Removed.
* Scripts/webkitpy/layout_tests/test_types/image_diff.py: Removed.
* Scripts/webkitpy/layout_tests/test_types/test_type_base.py: Removed.
* Scripts/webkitpy/layout_tests/test_types/test_type_base_unittest.py: Removed.
* Scripts/webkitpy/layout_tests/test_types/text_diff.py: Removed.
2011-02-25 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
Change results.json format to the one used by unexpected_results.json
https://bugs.webkit.org/show_bug.cgi?id=52267
Also add runtimes in milliseconds to the JSON and make the output format more compact.
Named the file full_results.json to avoid conflicting with the results.json
file the test-results server currently serves up.
* Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/printing.py:
* Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
* Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
2011-02-27 Adam Roben <aroben@apple.com>
Followup fix to r79827
* DumpRenderTree/win/UIDelegate.cpp:
(UIDelegate::webViewAddMessageToConsole): Replaced call to urlSuitableForTestResult with
lastPathComponent. This matches what we did here before r79827, and matches Mac.
2011-02-27 Dan Bernstein <mitz@apple.com>
Reviewed by Adam Roben.
Make bisect-builds print a trac link for the regression range
https://bugs.webkit.org/show_bug.cgi?id=55331
* Scripts/bisect-builds:
2011-02-27 Adam Roben <aroben@apple.com>
Use iswalpha instead of isalpha when dealing with wchar_ts in EditingDelegate
Fixes <http://webkit.org/b/55062> <rdar://problem/9059907> Crash beneath
EditingDelegate::checkSpellingOfString when running fast/forms/input-text-maxlength.html or
fast/forms/input-text-paste-maxlength.html on Windows with full page heap enabled
Reviewed by Anders Carlsson.
* DumpRenderTree/win/EditingDelegate.cpp:
(indexOfFirstWordCharacter):
(wordLength):
Use iswalpha instead of isalpha, since TCHAR is really just wchar_t.
2011-02-27 Adam Roben <aroben@apple.com>
Make ResourceLoadDelegate print URLs relative to the main resource on Windows
This matches Mac. (In fact, the code was ported from Mac!)
Fixes <http://webkit.org/b/55328> URLs printed by ResourceLoadDelegate on Windows don't
match Mac
Reviewed by Anders Carlsson.
* DumpRenderTree/win/DumpRenderTree.cpp:
(substringFromIndex): Added. Emulates -[NSString substringFromIndex:].
(urlSuitableForTestResult): Ported code from -[NSURL(DRTExtras)
_drt_descriptionSuitableForTestResult].
(cfStringRefToWString): Moved here from LayoutTestControllerWin.cpp.
* DumpRenderTree/win/DumpRenderTreeWin.h: Added declaration of cfStringRefToWString.
* DumpRenderTree/win/LayoutTestControllerWin.cpp: Moved cfStringRefToWString to
DumpRenderTree.cpp.
2011-02-27 Adam Roben <aroben@apple.com>
Change the order of Chang Shu's email addresses so the first one is their Bugzilla username
This makes Bugzilla autocompletion work correctly.
* Scripts/webkitpy/common/config/committers.py:
2011-02-27 Patrick Gansterer <paroga@webkit.org>
Reviewed by Andreas Kling.
[Qt] fast/loader/user-stylesheet-fast-path.html fails
https://bugs.webkit.org/show_bug.cgi?id=50911
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::setUserStyleSheetLocation):
Use QUrl::fromEncoded since the passed string is already encoded.
2011-02-27 Adam Roben <aroben@apple.com>
Ensure $testResultsDirectory is an absolute path before setting up Windows crash log saving
Windows 7 (and maybe Vista?) launches the post-mortem debugger with a working directory that
is different from old-run-webkit-test's. Because of this, the path to the crash log file
(which is relative to $testResultsDirectory) must be an absolute path. See the bug for more
details.
Fixes <http://webkit.org/b/55318> Crash logs aren't saved on Windows 7 when a relative
--results-directory path is used (like on the test slaves)
Reviewed by Dan Bernstein.
* Scripts/old-run-webkit-tests: Moved the call to setUpWindowsCrashLogSaving after we've
made $testResultsDirectory an absolute path.
2011-02-26 Dominic Mazzoni <dmazzoni@google.com>
Reviewed by Adam Barth.
m_dumpAccessibilityNotifications should be initialized.
https://bugs.webkit.org/show_bug.cgi?id=54874
* DumpRenderTree/chromium/AccessibilityController.cpp:
(AccessibilityController::AccessibilityController):
2011-02-26 Vsevolod Vlasov <vsevik@chromium.org>
Reviewed by Pavel Feldman.
DumpRenderTree should reset frame opener between tests.
https://bugs.webkit.org/show_bug.cgi?id=54874
Added clearOpener() calls between tests.
Some tests are using window.open() to change current window location.
This changes window.opener property whose nullity is checked before
showing xml viewer.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::resetTestController):
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues):
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
* DumpRenderTree/win/DumpRenderTree.cpp:
(resetWebViewToConsistentStateBeforeTesting):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::reset):
2011-02-26 Yael Aharon <yael.aharon@nokia.com>
Reviewed by Andreas Kling.
[Qt] Fix the focus in MiniBrowser.
https://bugs.webkit.org/show_bug.cgi?id=55288
Give focus to the web view when a load starts.
* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::urlChanged):
2011-02-26 Gyuyoung Kim <gyuyoung.kim@samsung.com>
Reviewed by Adam Barth.
[EFL] Add coding style exceptions for EFL port
https://bugs.webkit.org/show_bug.cgi?id=54733
The EFL APIs use EFL naming style, which includes both lower-cased and camel-cased,
underscore-sparated values. But, style checking rule have found parameter_name errors
in efl directories so far.
* Scripts/webkitpy/style/checker.py: Add prameter_name rule's exception and efl directories.
2011-02-26 David Levin <levin@chromium.org>
Reviewed by Darin Adler.
check-webkit-style should check for spaces after periods in comments
https://bugs.webkit.org/show_bug.cgi?id=55253
* Scripts/webkitpy/style/checkers/cpp.py: Added the check.
* Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests
and fixed the name of one related test.
2011-02-26 Kenneth Rohde Christiansen <kenneth@webkit.org>
Reviewed by Andreas Kling.
Make it possible to test the targetdensity-dpi support
https://bugs.webkit.org/show_bug.cgi?id=55142
Test the viewport meta tag feature targetdensity-dpi by
adding extra arguments to dumpConfigurationForViewport
Add the new arguments to the shared header and all
implementation.
* DumpRenderTree/LayoutTestController.cpp:
(dumpConfigurationForViewportCallback):
* DumpRenderTree/LayoutTestController.h:
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::dumpConfigurationForViewport):
* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::dumpConfigurationForViewport):
* DumpRenderTree/win/LayoutTestControllerWin.cpp:
(LayoutTestController::dumpConfigurationForViewport):
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::dumpConfigurationForViewport):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::dumpConfigurationForViewport):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
2011-02-26 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r79764.
http://trac.webkit.org/changeset/79764
https://bugs.webkit.org/show_bug.cgi?id=55295
"broke Chromium builds" (Requested by rniwa on #webkit).
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::resetTestController):
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues):
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
* DumpRenderTree/win/DumpRenderTree.cpp:
(resetWebViewToConsistentStateBeforeTesting):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::reset):
2011-02-26 Yongjun Zhang <yongjun_zhang@apple.com>
Reviewed by David Kilzer.
https://bugs.webkit.org/show_bug.cgi?id=48781
Add a resource load delegate method to query if WebCore should paint the default broken image for failed images.
Add a new resource load client method (shouldPaintBrokenImage). WebKit client can decide if WebCore
should paint the default broken image when an image fails to load or decode.
* DumpRenderTree/LayoutTestController.cpp:
(LayoutTestController::LayoutTestController):
(setShouldPaintBrokenImageCallback):
(LayoutTestController::staticFunctions):
(LayoutTestController::setShouldPaintBrokenImage):
* DumpRenderTree/LayoutTestController.h:
(LayoutTestController::shouldPaintBrokenImage):
* DumpRenderTree/mac/ResourceLoadDelegate.mm:
(-[ResourceLoadDelegate webView:shouldPaintBrokenImageForURL:]):
2011-02-26 Vsevolod Vlasov <vsevik@chromium.org>
Reviewed by Pavel Feldman.
DumpRenderTree should reset frame opener between tests.
https://bugs.webkit.org/show_bug.cgi?id=54874
Added clearOpener() calls between tests.
Some tests are using window.open() to change current window location.
This changes window.opener property whose nullity is checked before
showing xml viewer.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::resetTestController):
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues):
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
* DumpRenderTree/win/DumpRenderTree.cpp:
(resetWebViewToConsistentStateBeforeTesting):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::reset):
2011-02-25 Lukasz Slachciak <l.slachciak@samsung.com>
Reviewed by Kenneth Rohde Christiansen.
[EFL] Added EWebLauncher command line option to turn on/off frame flattening
https://bugs.webkit.org/show_bug.cgi?id=54809
* EWebLauncher/main.c: Handling -f option added.
(on_key_down):
(browserCreate):
(main):
2011-02-25 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r79324.
http://trac.webkit.org/changeset/79324
https://bugs.webkit.org/show_bug.cgi?id=55260
Leopard Debug is once again running out of address space while
linking (Requested by aroben on #webkit).
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2011-02-25 Tony Chang <tony@chromium.org>
Reviewed by James Robinson.
Clean up temp files left by crashing DRTs on chromium-win
https://bugs.webkit.org/show_bug.cgi?id=55182
* BuildSlaveSupport/build.webkit.org-config/master.cfg:
* BuildSlaveSupport/chromium/remove-crash-logs: Added.
2011-02-25 Adam Roben <aroben@apple.com>
Double-check that modifying the registry worked
On Windows Vista/7 with UAC enabled, regtool will fail to modify the registry, but it will
still return a successful exit code. So we double-check here that the value we tried to
write to the registry was really written.
Fixes <http://webkit.org/b/55227> old-run-webkit-tests falsely claims to be saving crash
logs on Windows Vista/7
Reviewed by Darin Adler.
* Scripts/old-run-webkit-tests:
(writeRegistryString): After trying to write the string value, check that it was really
written.
2011-02-25 Adam Roben <aroben@apple.com>
Set the 32-bit post-mortem debugger on 64-bit OSes
The post-mortem debugger is controlled by some registry values in
HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AeDebug. On 64-bit OSes, you can either
set the values in that key, which will be used for 64-bit apps, or in the equivalent key
under Wow6432Node, which will be used for 32-bit apps. Since DumpRenderTree and
WebKitTestRunner are 32-bit, we want to use the latter location.
Fixes <http://webkit.org/b/55225> old-run-webkit-tests sets the post-mortem debugger for
64-bit apps, but should set it for 32-bit apps instead
Reviewed by Darin Adler.
* Scripts/old-run-webkit-tests:
(readRegistryString):
(writeRegistryString):
Pass --wow32 to regtool so it will get/set values beneath Wow6432Node, which is where
Windows looks for the post-mortem debugger values when a 32-bit app crashes on a 64-bit OS.
Passing --wow32 on a 32-bit OS will not have any effect.
2011-02-25 Adam Roben <aroben@apple.com>
Clean up code to set up crash log saving on Windows
Fixes <http://webkit.org/b/55224> old-run-webkit-test's Windows registry code is ugly
Reviewed by David Kilzer.
* Scripts/old-run-webkit-tests:
(readRegistryString):
(writeRegistryString):
Moved code to read strings from and write strings to the registry here...
(setUpWindowsCrashLogSaving): ...from here.
2011-02-25 Adam Roben <aroben@apple.com>
Work around Cygwin's crash-suppression behavior
Cygwin calls ::SetErrorMode(SEM_FAILCRITICALERRORS), which any processes it launches will
inherit. This is bad for testing/debugging, as it causes the post-mortem debugger not to be
invoked. (Cygwin does this because it makes crashes more UNIX-y.) We reset the error mode
when our test apps launch to work around Cygwin's behavior.
Fixes <http://webkit.org/b/55222> Test apps crash silently (without invoking post-mortem
debugger) when launched from Cygwin 1.7
Reviewed by Darin Adler.
* DumpRenderTree/win/DumpRenderTree.cpp:
(main):
* TestWebKitAPI/win/main.cpp:
(main):
* WebKitAPITest/main.cpp:
(main):
* WebKitTestRunner/win/TestControllerWin.cpp:
(WTR::TestController::platformInitialize):
Call ::SetErrorMode(0) to undo Cygwin's folly.
2011-02-24 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>
Reviewed by Andreas Kling.
[Qt] Revert the support for QNAM affined to a different thread.
https://bugs.webkit.org/show_bug.cgi?id=55149
Qt 4.8 will have QNAM use its own thread internally by default,
no need to keep this complexity in WebKit.
This mainly reverts:
http://trac.webkit.org/changeset/73710
http://trac.webkit.org/changeset/73712
* QtTestBrowser/launcherwindow.cpp:
(LauncherWindow::~LauncherWindow):
(LauncherWindow::initializeView):
(LauncherWindow::createChrome):
* QtTestBrowser/launcherwindow.h:
(WindowOptions::WindowOptions):
* QtTestBrowser/webpage.cpp:
(WebPage::WebPage):
* QtTestBrowser/webpage.h:
2011-02-24 Adam Barth <abarth@webkit.org>
Reviewed by Eric Seidel.
Add a script for creating gypi files
https://bugs.webkit.org/show_bug.cgi?id=55113
This is a trivial script to create gypi files from the source code in
the current directory.
* Scripts/make-gypi: Added.
2011-02-24 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r79628.
http://trac.webkit.org/changeset/79628
https://bugs.webkit.org/show_bug.cgi?id=55195
"broke chromium win tests" (Requested by dpranke on #webkit).
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2011-02-24 Dirk Pranke <dpranke@chromium.org>
Reviewed by Mihai Parparita.
rebaseline-chromium-webkit-tests can't handle GPU variants
https://bugs.webkit.org/show_bug.cgi?id=55002
* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-02-24 Dirk Pranke <dpranke@chromium.org>
Reviewed by Mihai Parparita.
nrwt: log command used to start web server to --verbose
https://bugs.webkit.org/show_bug.cgi?id=55188
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
2011-02-24 Victoria Kirst <vrk@google.com>
Reviewed by Mihai Parparita.
[chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
https://bugs.webkit.org/show_bug.cgi?id=55089
This adds an alias into the media/ directory so that http/tests/media
tests can access the media resources when running in an httpd process.
* Scripts/webkitperl/httpd.pm:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
2011-02-24 Dirk Pranke <dpranke@chromium.org>
Reviewed by Mihai Parparita.
nrwt: enable multiple processes by default on Chromium Win.
https://bugs.webkit.org/show_bug.cgi?id=55163
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2011-02-24 Sam Weinig <sam@webkit.org>
Attempt to fix Qt.
* WebKitTestRunner/TestInvocation.cpp:
2011-02-24 James Robinson <jamesr@chromium.org>
Unreviewed, rolling out r79598.
http://trac.webkit.org/changeset/79598
https://bugs.webkit.org/show_bug.cgi?id=55089
Appears to be causing trouble with httpd on windows
* Scripts/webkitperl/httpd.pm:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
2011-02-24 Sam Weinig <sam@webkit.org>
Reviewed by Darin Adler.
Add basic pixel testing support to WebKitTestRunner
https://bugs.webkit.org/show_bug.cgi?id=55161
* Scripts/old-run-webkit-tests:
Don't print color space information when using WebKitTestRunner since
we don't need to change the screen.
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::initialize):
Add call to initializePlatformDefaults.
(WTR::InjectedBundle::done):
Return a dictionary on completion with both the text output and pixel dump override.
* WebKitTestRunner/InjectedBundle/InjectedBundle.h:
Add declaration for initializePlatformDefaults.
* WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
(WTR::LayoutTestController::LayoutTestController):
(WTR::LayoutTestController::dumpAsText):
* WebKitTestRunner/InjectedBundle/LayoutTestController.h:
(WTR::LayoutTestController::shouldDumpPixels):
Make setting dumpAsText override the dump pixels preference.
* WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: Added.
(WTR::InjectedBundle::initializePlatformDefaults):
Set platform defaults matching DumpRenderTree.
* WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp: Added.
(WTR::InjectedBundle::initializePlatformDefaults):
* WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp: Added.
(WTR::InjectedBundle::initializePlatformDefaults):
Add stubs for initializePlatformDefaults.
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::runTest):
(WTR::TestController::runTestingServerLoop):
(WTR::TestController::run):
Parse expected pixel results out of stdin and setup the invocation with it.
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::TestInvocation):
Initialize new members.
(WTR::TestInvocation::~TestInvocation):
(WTR::TestInvocation::setIsPixelTest):
(WTR::sizeWebViewForCurrentTest):
(WTR::TestInvocation::invoke):
(WTR::TestInvocation::dump):
(WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
* WebKitTestRunner/TestInvocation.h:
Dump pixels in addition to text by calling dumpPixelsAndCompareWithExpected.
* WebKitTestRunner/mac/PlatformWebViewMac.mm:
(WTR::PlatformWebView::PlatformWebView):
Use -[NSColorSpace genericRGBColorSpace] instead of the main screen color space.
* WebKitTestRunner/mac/TestInvocationMac.mm: Added.
(WTR::createCGContextFromPlatformView):
(WTR::computeMD5HashStringForContext):
(WTR::dumpBitmap):
(WTR::forceRepaintFunction):
(WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
Add mac specific pixel dumping support. Right now, we always pull pixels
from the window server.
* WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
* WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
* WebKitTestRunner/qt/WebKitTestRunner.pro:
* WebKitTestRunner/win/InjectedBundle.vcproj:
* WebKitTestRunner/win/WebKitTestRunner.vcproj:
Add new files.
2011-02-24 Chang Shu <cshu@webkit.org>
Reviewed by Kenneth Rohde Christiansen.
[Qt] Launch MiniBrowser if the test run is WebKit2.
https://bugs.webkit.org/show_bug.cgi?id=55145
* Scripts/old-run-webkit-tests:
* Scripts/run-launcher:
* Scripts/webkitdirs.pm:
2011-02-24 Victoria Kirst <vrk@google.com>
Reviewed by Mihai Parparita.
[chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
https://bugs.webkit.org/show_bug.cgi?id=55089
This adds an alias into the media/ directory so that http/tests/media
tests can access the media resources when running in an httpd process.
* Scripts/webkitperl/httpd.pm:
* Scripts/webkitpy/layout_tests/port/apache_http_server.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
2011-02-24 Andrew Wilson <atwilson@chromium.org>
Unreviewed, rolling out r79570.
http://trac.webkit.org/changeset/79570
https://bugs.webkit.org/show_bug.cgi?id=54874
Breaks chromium build because glue/mocks/mock_web_frame.h/cc
was not updated
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::resetTestController):
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues):
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
* DumpRenderTree/win/DumpRenderTree.cpp:
(resetWebViewToConsistentStateBeforeTesting):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::reset):
2011-02-24 Mahesh Kulkarni <mahesh.kulkarni@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] add android 2.2 user agent to QtTestBrowser
https://bugs.webkit.org/show_bug.cgi?id=55085
Credit to Forrest Hodgkins to find right user-agent which works for youtube.
* QtTestBrowser/useragentlist.txt:
2011-02-24 Adam Roben <aroben@apple.com>
Change FrameLoadDelegate to support any number of delegates with delayed work to process
This makes our behavior match Mac more closely, and allows us to remove an incorrect
assertion that was firing during some tests. (The assertion was claiming that there was
never more than one delegate with delayed work to process, but that was not the case.)
Fixes <http://webkit.org/b/55146> Assertion failure in FrameLoadDelegate::locationChangeDone
when running http/tests/navigation/back-twice-without-commit.html
Reviewed by Eric Carlson.
* DumpRenderTree/win/FrameLoadDelegate.cpp:
(delegatesWithDelayedWork): Added. Returns all FrameLoadDelegates that have delayed work to
process. A single delegate may appear in this Vector more than once (just as, on Mac, a
single delegate may have multiple performSelector requests).
(processWorkTimer): Pass the HWND to ::KillTimer, for pedantic brownie points. Added an
assertion that the timer firing is the shared process work timer. Instead of using the
single, global "delegate waiting for timer" delegate, give all delegates that have delayed
work to process a chance to process their work.
(FrameLoadDelegate::locationChangeDone): If we don't already have an active timer for
processing delayed work, create one. Then add ourselves to the delegatesWithDelayedWork
Vector so our processWork function will be called when the timer fires.
2011-02-24 Vsevolod Vlasov <vsevik@chromium.org>
Reviewed by Alexey Proskuryakov.
DumpRenderTree should reset frame opener between tests.
https://bugs.webkit.org/show_bug.cgi?id=54874
Added clearOpener() calls between tests.
Some tests are using window.open() to change current window location.
This changes window.opener property whose nullity is checked before
showing xml viewer.
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::resetTestController):
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues):
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetWebViewToConsistentStateBeforeTesting):
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
* DumpRenderTree/win/DumpRenderTree.cpp:
(resetWebViewToConsistentStateBeforeTesting):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::reset):
2011-02-24 Adam Roben <aroben@apple.com>
Don't trigger Windows builds for changes to WebCore.exp.in or the top-level GNUmakefile.am
Fixes <http://webkit.org/b/55144> Windows builders should not have built r79343 or r79440,
but did
Reviewed by Anders Carlsson.
* Scripts/webkitpy/common/config/build.py:
(_should_file_trigger_build): Loosened the pattern for GNUmakefile.am to include the
top-level GNUmakefile.am, and loosened the pattern for .exp files to include WebCore.exp.in.
* Scripts/webkitpy/common/config/build_unittest.py:
(ShouldBuildTest): Added some more tests.
2011-02-24 Andras Becsi <abecsi@webkit.org>
Reviewed by Laszlo Gombos.
[Qt] MinGW build fails to link
https://bugs.webkit.org/show_bug.cgi?id=55050
Prepend the libraries of subcomponents instead of appending them
to fix the library order according to the dependency of the libraries
* WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
* WebKitTestRunner/qt/WebKitTestRunner.pro:
2011-02-24 Gabor Loki <loki@webkit.org>
Reviewed by Csaba Osztrogonác.
[Qt] Remove ARMv5 Release buildbot.
Remove ARMv5 Release buildbot, because there is very little difference
between ARMv5 and ARMv7 Release bots, and there is also very little
interest in ARMv5 build slave (compared to ARMv7).
* BuildSlaveSupport/build.webkit.org-config/config.json:
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2011-02-22 Ojan Vafai <ojan@chromium.org>
Reviewed by Tony Chang.
fix revision number finding when creating git patches
https://bugs.webkit.org/show_bug.cgi?id=55024
HEAD~n does not walk merge commits correctly. git log however does.
* Scripts/webkitpy/common/checkout/scm.py:
* Scripts/webkitpy/common/checkout/scm_unittest.py:
2011-02-23 Kenneth Russell <kbr@google.com>
Unreviewed, rolling out r79387.
https://bugs.webkit.org/show_bug.cgi?id=54885
Reapply the code changes from r79038 since they weren't the root
cause of the test timeouts.
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::reset):
2011-02-23 Siddharth Mathur <siddharth.mathur@nokia.com>
Reviewed by Laszlo Gombos.
[Qt] Make sure Symbian binary UIDs are unique
* MiniBrowser/qt/MiniBrowser.pro: Resolve the UID collision with
QtTestBrowser.pro by changing the UID.
2011-02-23 Patrick Gansterer <paroga@webkit.org>
Reviewed by Darin Adler.
Rename PLATFORM(CF) to USE(CF)
https://bugs.webkit.org/show_bug.cgi?id=53540
* DumpRenderTree/config.h:
2011-02-23 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
nrwt: turn on multiprocessing on chromium_linux, clean up unit tests for worker model defaults
https://bugs.webkit.org/show_bug.cgi?id=55016
* Scripts/webkitpy/layout_tests/port/base.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux.py:
* Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py: Added.
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
2011-02-23 Tony Chang <tony@chromium.org>
Reviewed by Kent Tamura.
[chromium] DumpRenderTree --check-layout-test-sys-deps leaks 4 temp dirs
https://bugs.webkit.org/show_bug.cgi?id=55004
* DumpRenderTree/chromium/DumpRenderTree.cpp:
(WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
(WebKitSupportTestEnvironment::~WebKitSupportTestEnvironment):
(main): Use a C++ object to ensure that TearDownTestEnvironment
gets called when --check-layout-test-sys-deps is called
2011-02-23 Gopal Raghavan <gopal.1.raghavan@nokia.com>
Reviewed by Anders Carlsson.
[Qt] [Qt] check webkit style fails in qt/tests
https://bugs.webkit.org/show_bug.cgi?id=54998
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.
* Scripts/webkitpy/style/checker.py:
2011-02-22 Vsevolod Vlasov <vsevik@chromium.org>
Reviewed by Pavel Feldman.
DumpRenderTree should dump as text if test path contains "dumpAsText/"
https://bugs.webkit.org/show_bug.cgi?id=54873
* DumpRenderTree/chromium/LayoutTestController.h:
(LayoutTestController::setShouldDumpAsText):
(LayoutTestController::setShouldGeneratePixelResults):
* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::runFileTest):
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(shouldDumpAsText):
(dump):
(runTest):
* DumpRenderTree/mac/DumpRenderTree.mm:
(shouldDumpAsText):
(runTest):
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::isDumpAsTextTest):
(WebCore::DumpRenderTree::open):
* DumpRenderTree/win/DumpRenderTree.cpp:
(shouldDumpAsText):
(runTest):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::dump):
2011-02-22 Kenneth Russell <kbr@google.com>
Reviewed by James Robinson.
Many tests started timing out on the chrome mac canaries around r79035
https://bugs.webkit.org/show_bug.cgi?id=54885
Revert the code changes, but not the Chromium DEPS roll, from
r79038 to see whether this clears up the timing-out tests. Was not
able to reproduce the timeouts locally with a Release mode build
of DRT (on Snow Leopard).
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::reset):
2011-02-22 Dirk Pranke <dpranke@chromium.org>
Reviewed by James Robinson.
nrwt: --platform chromium-gpu doesn't work with --worker-model=processes
https://bugs.webkit.org/show_bug.cgi?id=55009
* Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2011-02-18 Mihai Parparita <mihaip@chromium.org>
Reviewed by Ojan Vafai.
Default to 500 for --exit-after-n-failures
https://bugs.webkit.org/show_bug.cgi?id=54773
Make 500 the default value for --exit-after-n-failures (ORWT bots
already default to this as of r75726). Similarly, change the default for
--exit-after-n-crashes-or-timeouts to 20.
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2011-02-02 Martin Robinson <mrobinson@igalia.com>
Reviewed by Xan Lopez.
[GTK] DRT needs an implementation of LayoutTestController.setSelectTrailingWhitespace
https://bugs.webkit.org/show_bug.cgi?id=53603
Add support for LayoutTestController.setSelectTrailingWhitespace by calling through
to DumpRenderTreeSupportGtk for this functionality.
* DumpRenderTree/gtk/DumpRenderTree.cpp:
(resetDefaultsToConsistentValues): Always call LayoutTestController.setSelectTrailingWhitespace to
false when reseting values between tests.
* DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
(LayoutTestController::setSelectTrailingWhitespaceEnabled): Call LayoutTestController.setSelectTrailingWhitespace
with the appropriate value.
2011-02-22 Adam Roben <aroben@apple.com>
Don't trigger a Windows build when the top-level ChangeLog changes
r79320 is an example of a change that shouldn't have built on Windows, but did, because of
this bug.
Fixes <http://webkit.org/b/54957> Windows bots build when top-level ChangeLog changes, but
shouldn't
Reviewed by Anders Carlsson.
* Scripts/webkitpy/common/config/build.py:
(_should_file_trigger_build): Modified the ChangeLog pattern to match the top-level
ChangeLog, and re-sorted it in the list.
* Scripts/webkitpy/common/config/build_unittest.py:
(ShouldBuildTest): Added a test that ChangeLogs don't trigger builds.
2011-02-22 Sheriff Bot <webkit.review.bot@gmail.com>
Unreviewed, rolling out r78691.
http://trac.webkit.org/changeset/78691
https://bugs.webkit.org/show_bug.cgi?id=54956
Leopard Debug is successfully building again (Requested by
aroben on #webkit).
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2011-02-22 Ilya Tikhonovsky <loislo@chromium.org>
Reviewed by Pavel Feldman.
Web Inspector: flakyness of inspector tests.
https://bugs.webkit.org/show_bug.cgi?id=54729
As far as we have the protocol with sequence numbers we can simplify test harness support
and drop out chomium specific methods.
* DumpRenderTree/DumpRenderTree.gypi:
* DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
(DRTDevToolsAgent::sendMessageToInspectorFrontend):
(DRTDevToolsAgent::runtimePropertyChanged):
(DRTDevToolsAgent::asyncCall):
(DRTDevToolsAgent::call):
* DumpRenderTree/chromium/DRTDevToolsAgent.h:
* DumpRenderTree/chromium/DRTDevToolsCallArgs.cpp: Removed.
* DumpRenderTree/chromium/DRTDevToolsCallArgs.h: Removed.
* DumpRenderTree/chromium/DRTDevToolsClient.cpp:
(DRTDevToolsClient::sendMessageToBackend):
(DRTDevToolsClient::asyncCall):
(DRTDevToolsClient::call):
* DumpRenderTree/chromium/DRTDevToolsClient.h:
2011-02-22 Andras Becsi <abecsi@webkit.org>
Reviewed by Laszlo Gombos.
[Qt] Redesign the build system
https://bugs.webkit.org/show_bug.cgi?id=51339
Part 2.
Build WebCore as a static library, compile the WebKit API and WebKit2 API
in a final step and link to WebKit2, WebCore and JSC libraries to fix
linking issues resulting from stripped away symbols.
* WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Remove superfluous includepaths.
2011-02-21 Lukasz Slachciak <l.slachciak@samsung.com>
Reviewed by Antonio Gomes.
[EFL] Restored spatial navigation option in EWebLauncher since #18662 bug is closed
https://bugs.webkit.org/show_bug.cgi?id=54806
* EWebLauncher/main.c:
(on_key_down): calling ewk_view_setting_spatial_navigation_set when "F12" pressed added.
2011-02-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Kent Tamura.
nrwt: revert behavior on 10.5 to "old-threads". It's possible
that r79062 has introduced a bunch of flakiness, although I
can't explain why right now. This change should restore the
old behavior, and then we can look at the build logs to see
if the tests stop timing out or being flaky.
https://bugs.webkit.org/show_bug.cgi?id=54925
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2011-02-21 Dirk Pranke <dpranke@chromium.org>
Reviewed by Dimitri Glazkov.
nrwt: fix typo causing us to try and repeatedly shut down the servers
https://bugs.webkit.org/show_bug.cgi?id=54904
* Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
2011-02-21 Alejandro G. Castro <alex@igalia.com>
Reviewed by Csaba Osztrogonác.
REGRESSION(r79157):
plugins/get-url-with-javascript-destroying-plugin.html make the
next test crash
https://bugs.webkit.org/show_bug.cgi?id=54863
* GNUmakefile.am:
2011-02-20 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
test-webkitpy: stop skipping a bunch of tests on win32
https://bugs.webkit.org/show_bug.cgi?id=54788
* Scripts/webkitpy/common/net/testoutputset_unittest.py:
* Scripts/webkitpy/test/main.py:
2011-02-20 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
Fix various bugs related to handling of dir separators and real
vs. fake filesystems that were causing test-webkitpy to fail
tests on win32. These bugs have a variety of causes but are
lumped together because they should be easy to review.
https://bugs.webkit.org/show_bug.cgi?id=54700
* Scripts/webkitpy/common/net/testoutput.py:
Here we were using os.path.sep but not escaping it prior to
using in a regexp, which caused bad things on win32.
* Scripts/webkitpy/common/net/testoutput_unittest.py:
Same here.
* Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
Here we were using the native filesystem instead of the mock
filesystem, making some tests unnecessarily platform dependent.
Also we weren't setting the default configuration for the
TestChromiumWinPort.
* Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py:
Here we *shouldn't* use os.path.join(), just so the code becomes
less platform dependent.
* Scripts/webkitpy/style/checker.py:
Here we should be using os.path.sep instead of "/"; the tests
require the native filesystem.
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
Here the underlying code is using os.path.sep, so we probably
should as well.
* Scripts/webkitpy/style/test_expectations_unittest.py:
Here we should be using the filesystem hanging of the port
object (which is the mocked filesystem) rather than the native
one.
2011-02-20 Anders Carlsson <andersca@apple.com>
Reviewed by Maciej Stachowiak.
Crash when a plug-in requests a javascript: url that destroys the plug-in
https://bugs.webkit.org/show_bug.cgi?id=54837
<rdar://problem/9005475>
Add new plug-in test.
* DumpRenderTree/DumpRenderTree.gypi:
* DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
* DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
(PluginTest::NPN_GetURL):
* DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
* DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURLDestroyingPlugin.cpp: Added.
(GetURLWithJavaScriptURLDestroyingPlugin::GetURLWithJavaScriptURLDestroyingPlugin):
(GetURLWithJavaScriptURLDestroyingPlugin::NPP_New):
* DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
2011-02-19 Patrick Gansterer <paroga@webkit.org>
Reviewed by Andreas Kling.
Make WinCE a core builder
https://bugs.webkit.org/show_bug.cgi?id=54801
* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2011-02-19 Lukasz Slachciak <l.slachciak@samsung.com>
Reviewed by Andreas Kling.
Fixed problem with launching EWebLauncher with "run-launcher --efl". Incorrect path referring to non-existing .libs directory ws set.
https://webkit.org/b/54778
* Scripts/webkitdirs.pm:
2011-02-18 David Levin <levin@chromium.org>
Reviewed by Eric Seidel.
check-webkit-style: Misses brace style error when the line with the { has a } in it.
https://bugs.webkit.org/show_bug.cgi?id=54769
* Scripts/webkitpy/style/checkers/cpp.py: Fixed the check for the close brace
to only look after the last open brace, so that the open brace in this line
"} else {" will still be able to trigger the error.
* Scripts/webkitpy/style/checkers/cpp_unittest.py: Added related unit tests.
2011-02-18 Zan Dobersek <zandobersek@gmail.com>
Reviewed by Martin Robinson.
[Gtk] Re-enable meter tag support
https://bugs.webkit.org/show_bug.cgi?id=54762
Meter tag is supported on the Gtk port, enable it by default.
* Scripts/build-webkit:
2011-02-18 David Levin <levin@chromium.org>
Reviewed by Eric Seidel.
check-webkit-style falsely complains about WebKitGTK+ public headers
https://bugs.webkit.org/show_bug.cgi?id=54650
* Scripts/webkitpy/style/checker.py: Add some exceptions for the gtk files.
* Scripts/webkitpy/style/checkers/cpp.py: Restrict the checks for WEBKIT_API
to the chromium directory (and improved the checks slightly).
* Scripts/webkitpy/style/checkers/cpp_unittest.py: Added corresponding tests.
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by James Robinson.
new-run-webkit-tests: enable multiple processes by default on
mac. This change removes the artificial restrictions we placed
on the # of child processes we used with the old threading
model, and switches to the new message-based model and multiple
processes, where available. If multiple processes are not
available (Leopard / Python 2.5), then we use the 'inline' model
instead of the 'old-threads' model on the 'mac' port or one
process on the 'chromium-mac' port. We need additional testing
to see if the new 'threads' model works reliably and is worth
supporting, or if we should just live with things being slightly
slow.
https://bugs.webkit.org/show_bug.cgi?id=54596
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-02-18 Andrew Wilson <atwilson@chromium.org>
Unreviewed, rolling out r79047.
http://trac.webkit.org/changeset/79047
https://bugs.webkit.org/show_bug.cgi?id=54596
Broke canary bots - please watch the canaries the next time
you land this
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
2011-02-18 Emil A Eklund <eae@chromium.org>
Reviewed by Darin Adler.
Crash in EventHandler::sendContextMenuEventForKey
https://bugs.webkit.org/show_bug.cgi?id=54495
Add support for the context menu key (VK_APPS) to EventSender for
platforms that support the key.
* DumpRenderTree/chromium/EventSender.cpp:
(EventSender::keyDown):
* DumpRenderTree/gtk/EventSender.cpp:
(keyDownCallback):
* DumpRenderTree/qt/EventSenderQt.cpp:
(EventSender::keyDown):
* DumpRenderTree/win/EventSender.cpp:
(keyDownCallback):
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
new-run-webkit-tests: enable multiple processes by default on mac
https://bugs.webkit.org/show_bug.cgi?id=54596
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/mac.py:
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
Skip pretty patch unit tests if ruby isn't installed.
https://bugs.webkit.org/show_bug.cgi?id=54699
* Scripts/webkitpy/common/prettypatch_unittest.py:
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
Skip mac.test_skipped_file_paths() when running on win32; the
test has hardcoded '/' directory separators and since it's a
port-specific test and we have coverage on other ports it's not
worth it to make the test more generic.
https://bugs.webkit.org/show_bug.cgi?id=54698
* Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
Fix webkitpy.layout_tests.port.config_unittest to work on Win32.
https://bugs.webkit.org/show_bug.cgi?id=54703
* Scripts/webkitpy/layout_tests/port/config_unittest.py:
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
Change the default port we're mocking out on win32 when we
run mock_drt_unittest. Normally we'd default to the
port-specific default, but that would be "win", which doesn't
currently work.
https://bugs.webkit.org/show_bug.cgi?id=54708
* Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2011-02-18 Kenneth Russell <kbr@google.com>
Reviewed by James Robinson.
[chromium] Clean up initialization of minimum timer interval
https://bugs.webkit.org/show_bug.cgi?id=54772
Use new GetForegroundTabTimerInterval function in webkit_support
to reset the minimum timer interval before each test.
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::reset):
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
r78494 fixed the chromium_win unittest failing under win32,
but not cygwin. This fixes cygwin as well.
https://bugs.webkit.org/show_bug.cgi?id=54652
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
2011-02-18 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai, Mihai Parparita.
new-run-webkit-tests: r78522 made it impossible to create a
'chromium-win' or 'chromium-mac' port; they would always get the
version tacked on to the end. It turns out that we actually need
to be able to create these ports in order for
rebaseline-chromium-webkit-tests to work correctly given the
broken coverage of our layout bots. As soon as we get SL and
Win 7 bots running reliably on the canaries, we can get rid of
this change. Note that we only need to change the chromium_mac
and win ports because linux doesn't yet use versions and that
script doesn't work with non-chromium ports.
Also, this change fixes a bug where we were using
"platform/chromium-win-win7" and
"platform/chromium-mac-snowleopard" as the baseline dirs for
Win 7 and SL; we should've been using the generic directories
instead since they are the newest versions of the os's.
https://bugs.webkit.org/show_bug.cgi?id=54691
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
* Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
* Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
2011-02-16 MORITA Hajime <morrita@google.com>
Reviewed by Kent Tamura.
[Chromium][DRT] some PostDelayedTask() accesses possibly dangling this pointer.
http://webkit.org/b/53899
Replaced static invocation wrappers with MethodTask objects,
which are removed on WebViewHost destructor.
* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::requestCheckingOfText):
(WebViewHost::scheduleAnimation):
(WebViewHost::closeWidgetSoon):
* DumpRenderTree/chromium/WebViewHost.h:
(WebViewHost::HostMethodTask::HostMethodTask):
(WebViewHost::HostMethodTask::runIfValid):
(WebViewHost::taskList):
2011-02-07 Hayato Ito <hayato@chromium.org>
Reviewed by Ojan Vafai.
Get rid of code which writes test results from test_type's
compare_output() method.
Writing test results now happens in single_test_runner, using
newly introduced test_result_writer module.
Subsequent patches will eliminate test_type/* classes.
https://bugs.webkit.org/show_bug.cgi?id=53908
* Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
* Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py: Added.
* Scripts/webkitpy/layout_tests/test_types/image_diff.py:
* Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
* Scripts/webkitpy/layout_tests/test_types/text_diff.py:
2011-02-17 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
Add blacklist of remaining webkitpy/* modules that fail on
Win32, so that test-webkitpy will at least run cleanly.
https://bugs.webkit.org/show_bug.cgi?id=54709
* Scripts/webkitpy/layout_tests/test/main.py:
2011-02-17 Simon Fraser <simon.fraser@apple.com>
Reviewed by Sam Weinig.
Set a preference to ensure that scrollbars are in a predictable
state for DRT.
* DumpRenderTree/mac/DumpRenderTree.mm:
(resetDefaultsToConsistentValues):
2011-02-17 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
test-webkitpy doesn't work with the multiprocessing module on win32
or cygwin, so we skip it for now.
https://bugs.webkit.org/show_bug.cgi?id=54520
* Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
* Scripts/webkitpy/layout_tests/layout_package/run_webkit_tests_unittest.py:
2011-02-17 Dirk Pranke <dpranke@chromium.org>
Reviewed by Adam Roben.
Fix version, baseline_search_path for chromium-win-win7. Also
fix chromium_win_unittests, and fix port_testcase.make_port()
to correctly use the options passed into it and only set
results_directory() if necessary.
https://bugs.webkit.org/show_bug.cgi?id=54652
* Scripts/webkitpy/layout_tests/port/chromium_win.py:
* Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
2011-02-17 Dirk Pranke <dpranke@chromium.org>
Reviewed by Ojan Vafai.
Fix a hang in new-run-webkit-tests when run with
--dry-run --worker-model=processes
https://bugs.webkit.org/show_bug.cgi?id=54595
* Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2011-02-17 Kevin Ollivier <kevino@theolliviers.com>
[wx] Build fixes after recent changes.
* DumpRenderTree/wx/LayoutTestControllerWx.cpp:
(LayoutTestController::setMinimumTimerInterval):
* wx/build/settings.py:
2011-02-17 Csaba Osztrogonác <ossy@webkit.org>
Reviewed by Andreas Kling.
[Qt] ImageDiff sometimes hangs
https://bugs.webkit.org/show_bug.cgi?id=54641
* DumpRenderTree/qt/ImageDiff.cpp: Clear actualImage and baselineImage after all comparison.
(main):
2011-02-17 Peter Varga <pvarga@webkit.org>
Reviewed by Csaba Osztrogonác.
The run-sunspider --v8-suite doesn't work
https://bugs.webkit.org/show_bug.cgi?id=54660
* Scripts/run-sunspider:
2011-02-17 Gabor Rapcsanyi <rgabor@webkit.org>
Reviewed by Csaba Osztrogonác.
[NRWT] Fix environment setup on Qt port.
https://bugs.webkit.org/show_bug.cgi?id=54549
* Scripts/webkitpy/layout_tests/port/qt.py:
2011-02-16 Yael Aharon <yael.aharon@nokia.com>
Reviewed by Andreas Kling.
[Qt] Enable tiled backing store by default in MiniBrowser.
https://bugs.webkit.org/show_bug.cgi?id=54581
Instead of a flag to turn it on, a new flag is defined to turn
tiled backing store off.
* MiniBrowser/qt/MiniBrowserApplication.cpp:
(MiniBrowserApplication::handleUserOptions):
* MiniBrowser/qt/MiniBrowserApplication.h:
(WindowOptions::WindowOptions):
2011-02-16 Eric Seidel <eric@webkit.org>
Reviewed by Mihai Parparita.
flaky test archive uploads are always 22 bytes long
https://bugs.webkit.org/show_bug.cgi?id=54593
Speculative fix for the 22-bytes bug.
* Scripts/webkitpy/tool/bot/flakytestreporter.py:
2011-02-16 Anna Cavender <annacc@chromium.org>
Reviewed by Mihai Parparita.
http/tests/media should be served over HTTP (not from local file)
https://bugs.webkit.org/show_bug.cgi?id=54028
Removed special file:// treatment of http/tests/media
* Scripts/old-run-webkit-tests:
* Scripts/webkitpy/layout_tests/port/base.py:
2011-02-16 Mihai Parparita <mihaip@chromium.org>
Reviewed by Ojan Vafai.
ChangeLog.parse_latest_entry_from_file does not handle rolled over ChangeLogs
https://bugs.webkit.org/show_bug.cgi?id=54609
r78737 rolled over ChangeLogs, and webkit-patch complains when landing
patches such as this one that are the first to add an entry to the
ChangeLog (since they don't see a second date line to know where the
entry ends).
* Scripts/webkitpy/common/checkout/changelog.py:
* Scripts/webkitpy/common/checkout/changelog_unittest.py:
== Rolled over to ChangeLog-2011-02-16 ==