Timeline


and

11/29/09:

20:50 Changeset [51481] by bfulgham@webkit.org

Build fix.

Correct draw signature used in ImageCairoWin.cpp.

  • platform/graphics/win/ImageCairoWin.cpp:

(WebCore::BitmapImage::getHBITMAPOfSize):
(WebCore::BitmapImage::drawFrameMatchingSourceSize):

20:15 Changeset [51480] by oliver@apple.com

postMessage should serialize File objects
https://bugs.webkit.org/show_bug.cgi?id=31955

Reviewed by Sam Weinig.

Update SerializedScriptValue to include support for
File objects in the serialized object graph.

19:54 Changeset [51479] by eric@webkit.org

2009-11-29 Eric Seidel <eric@webkit.org>

Add Erik Arvidsson to committers.py since by r51326 he clear has commit rights.

  • Scripts/modules/committers.py:
19:45 Changeset [51478] by simon.fraser@apple.com

2009-11-29 Simon Fraser <simon.fraser@apple.com>

Rubber-stamped by Eric Seidel.

Fix up some compositing layout tests.

  • compositing/layers-inside-overflow-scroll.html:
  • compositing/self-painting-layers.html: Use the 'loadeddata' event, instead of 'load', to know when the video loads.
  • platform/mac/compositing/repaint/content-into-overflow-expected.checksum:
  • platform/mac/compositing/repaint/content-into-overflow-expected.png:
  • platform/mac/compositing/repaint/overflow-into-content-expected.checksum:
  • platform/mac/compositing/repaint/overflow-into-content-expected.png:
  • platform/mac/compositing/self-painting-layers-expected.checksum:
  • platform/mac/compositing/self-painting-layers-expected.png: Update with correct pixel results.
19:26 Changeset [51477] by eric@webkit.org

2009-11-29 Eric Seidel <eric@webkit.org>

No review, just adding missing svn property.

Set svn:executable property on new scripts.
These were added by commit-queue commits, but the executable
property was lost due to a bug in svn-apply:
https://bugs.webkit.org/show_bug.cgi?id=27204

  • Scripts/update-webkit-chromium: Added property svn:executable.
  • Scripts/validate-committer-lists: Added property svn:executable.
19:21 Changeset [51476] by simon.fraser@apple.com

2009-11-29 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein.

Optimize the hierarchy rebuilding of compositing layers
https://bugs.webkit.org/show_bug.cgi?id=31879

When updating the compositing layer hierarchy, instead of removing all
child layers and then re-adding them one by one, build a vector of child
layers, and set them in one go.

  • platform/graphics/GraphicsLayer.h:
  • platform/graphics/GraphicsLayer.cpp: (WebCore::GraphicsLayer::setChildren): New method that takes a Vector of child GraphicsLayers.
  • platform/graphics/mac/GraphicsLayerCA.h: Override setChildren().
  • platform/graphics/win/GraphicsLayerCACF.h: Ditto.
  • platform/graphics/mac/GraphicsLayerCA.mm: (WebCore::GraphicsLayerCA::setChildren): Implement setChildren() to set the bit that notes that sublayers changed.
  • platform/graphics/win/GraphicsLayerCACF.cpp: (WebCore::GraphicsLayerCACF::setChildren): Implement setChildren() to udpate sublayers, with a note that this is not efficient.
  • rendering/RenderLayerCompositor.h:
  • rendering/RenderLayerCompositor.cpp: (WebCore::RenderLayerCompositor::updateCompositingLayers): When updating compositing layers, use the faster updateLayerTreeGeometry() if we know that no layer hierarchy changes are needed, and, at the root, use the vector returned from rebuildCompositingLayerTree() to attach the root layer. (WebCore::RenderLayerCompositor::rebuildCompositingLayerTree): Changed to collect child layers into Vectors of GraphicsLayers, which can be set as layer children in one go. (WebCore::RenderLayerCompositor::updateLayerTreeGeometry): Split out from rebuildCompositingLayerTree() for simplicity, and called when we just need to update layer geometry, without doing any reparenting.
19:12 Changeset [51475] by eric@webkit.org

2009-11-29 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Need a way to validate that committers.py includes all committers/reviewers
https://bugs.webkit.org/show_bug.cgi?id=30970

Add a script which knows how to compare our public committer/reviewer
lists and show discrepancies between them.

Validates mailing lists vs. committers.py:

  • committers.py committers missing from webkit-committers@lists
  • webkit-committers@lists members missing from committers.py
  • committers.py reviewers missing from webkit-reviewers@lists
  • webkit-reviewers@lists members missing from committers.py
  • webkit-reviewers@lists members missing from committers.py reviewer list

Validates committers.py vs. trunk/ SVN history:

  • committers who have not committed in over a year
  • SVN committers missing from committers.py
  • committers.py members who have no record in SVN.

All of these lists still show "false positives" until a few more committers.py updates are made, like:
https://bugs.webkit.org/show_bug.cgi?id=31366

  • Scripts/validate-committer-lists: Added.
19:10 Changeset [51474] by dbates@webkit.org

2009-11-29 Daniel Bates <dbates@webkit.org>

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=31971

Updated comment to reflect latest XSSAuditor bindings.

No functionality was changed. So, no new tests.

  • page/XSSAuditor.h:
18:45 Changeset [51473] by dbates@webkit.org

2009-11-29 Daniel Bates <dbates@webkit.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=31969

Removes unnecessary #include files.

No functionality was changed. So, no new tests.

  • css/CSSComputedStyleDeclaration.cpp: Removed include CachedImage.h, and Pair.h
  • css/CSSCursorImageValue.cpp: Removed include RenderStyle.h
  • css/CSSFontFaceSrcValue.cpp: Removed include Node.h
  • css/CSSFontSelector.cpp: Removed include NodeList.h
  • css/CSSGradientValue.cpp: Removed include GraphicsContext.h, ImageBuffer.h
  • css/CSSImageValue.cpp: Removed include RenderStyle.h
  • css/CSSImportRule.cpp: Removed include MediaList.h
  • css/CSSMutableStyleDeclaration.cpp: Removed include CSSProperty.h
  • css/CSSRule.cpp: Removed include CSSStyleSheet.h
  • css/CSSStyleSelector.cpp: Removed include CSSFontFace.h, CSSFontFaceSource.h, and CSSProperty.h
  • page/EventSource.h: Removed include EventListener.h
17:32 Changeset [51472] by dbates@webkit.org

2009-11-29 Daniel Bates <dbates@webkit.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=31965

Removed #include KeyframeList.h from CSSStyleSelector.h. Instead,
forward declared it.

As a side effect, we need to #include KeyframeList.h in
RenderLayerBacking.cpp.

No functionality was changed. So, no new tests.

  • css/CSSStyleSelector.cpp:
  • css/CSSStyleSelector.h: Removed #include KeyframeList.h
  • rendering/RenderLayerBacking.cpp: Added #include KeyframeList.h
17:02 Changeset [51471] by eric@webkit.org

2009-11-29 Shu Chang <Chang.Shu@nokia.com>

Reviewed by Eric Seidel.

Test using Id as index of node list.
https://bugs.webkit.org/show_bug.cgi?id=31428

  • fast/dom/Element/id-in-node-list-index01-expected.txt: Added.
  • fast/dom/Element/id-in-node-list-index01.html: Added.

2009-11-29 Shu Chang <Chang.Shu@nokia.com>

Reviewed by Eric Seidel.

Continue to search for matching node in the case where multiple nodes
have the same id.
https://bugs.webkit.org/show_bug.cgi?id=31428

Test: fast/dom/Element/id-in-node-list-index01.html

  • dom/DynamicNodeList.cpp: (WebCore::DynamicNodeList::itemWithName):
16:50 Changeset [51470] by dbates@webkit.org

2009-11-29 Daniel Bates <dbates@webkit.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=31966

Removed #include RenderStyle.h from file KeyframeAnimation.h. Instead,
forward declared it.

No functionality was changed. So, no new tests.

  • page/animation/KeyframeAnimation.cpp:
  • page/animation/KeyframeAnimation.h:
16:06 Changeset [51469] by dbates@webkit.org

2009-11-29 Daniel Bates <dbates@webkit.org>

Unreviewed, fix change log entry date.

For some reason, bugzilla-tool did not update the date in the change
log entry for my last commit (r51468). So, this commit fixes the date
of that entry.

15:55 Changeset [51468] by dbates@webkit.org

2009-11-28 Daniel Bates <dbates@webkit.org>

Reviewed by Dan Bernstein.

https://bugs.webkit.org/show_bug.cgi?id=31964

Removed #include StringImpl, StyleImage. Instead, forward declared them.
Also, we can substitute #include <wtf/PassRefPtr.h> for #include PlatformString.h,
since it seems we only really used it to include PassRefPtr.h.

No functionality was changed. So, no new tests.

  • rendering/style/ContentData.h:
12:11 Changeset [51467] by eric@webkit.org

2009-11-29 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Eric Seidel.

[Qt] Mac build has JIT disabled
https://bugs.webkit.org/show_bug.cgi?id=31828

  • wtf/Platform.h: Enable JIT for Qt Mac builds
07:57 Changeset [51466] by eric@webkit.org

2009-11-29 Zoltan Horvath <zoltan@webkit.org>

Reviewed by Eric Seidel.

Allow custom memory allocation control for classes of the rendering and storage directory in WebCore
https://bugs.webkit.org/show_bug.cgi?id=31906

Inherits the following classes from Noncopyable because these are instantiated
by 'new' and no need to be copyable:

class/struct name - instantiated at: WebCore/'location'

class SQLTransactionClient - storage/DatabaseThread.cpp:45
class SQLTransactionCoordinator - storage/DatabaseThread.cpp:46
class OriginUsageRecord - storage/OriginQuotaManager.cpp:66
class DatabaseTracker - storage/DatabaseTracker.cpp:62
class ScrollbarTheme - (its child class) rendering/RenderScrollbarTheme.cpp:35
class RenderSelectionInfoBase - (its child class) rendering/RenderView.cpp:310
class RenderOverflow - rendering/RenderBox.cpp:2846

Inherits the following classes from FastAllocBase because these are instantiated by 'new':

struct FillLayer - css/CSSStyleSelector.cpp:197
struct ShadowData - rendering/style/ShadowData.cpp:35
class CounterContent - css/CSSStyleSelector.cpp:4111

  • platform/ScrollbarTheme.h:
  • rendering/RenderOverflow.h:
  • rendering/RenderSelectionInfo.h:
  • rendering/style/CounterContent.h:
  • rendering/style/FillLayer.h:
  • rendering/style/ShadowData.h:
  • storage/DatabaseTracker.h:
  • storage/OriginUsageRecord.h:
  • storage/SQLTransactionClient.h:
  • storage/SQLTransactionCoordinator.h:

11/28/09:

22:36 QtWebKitBackportingFixes edited by laszlo.1.gombos@nokia.com
(diff)
22:31 QtWebKitBackportingFixes edited by laszlo.1.gombos@nokia.com
(diff)
17:53 Changeset [51465] by abarth@webkit.org

2009-11-28 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] style-queue sends ~100 requests to QueueStatusServer every 5 minutes
https://bugs.webkit.org/show_bug.cgi?id=31950

Now we cache the last status that we get back from QueueStatusServer.
Eventually we'll have to do something more fancy if we want to support
a "try again" button on QueueStatusServer, but we can cross that bridge
when we come to it.

  • Scripts/modules/patchcollection.py:
17:50 Changeset [51464] by abarth@webkit.org

2009-11-28 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

StyleQueue scans ~100 bug pages every 5 minutes
https://bugs.webkit.org/show_bug.cgi?id=31947

Instead of getting the pending-review attachment ids by scanning each
bug (which results in a network request), we should just get them all
from webkit.org/pending-review in one shot.

  • Scripts/modules/bugzilla.py:
  • Scripts/modules/bugzilla_unittest.py:
  • Scripts/modules/commands/queries.py:
  • Scripts/modules/commands/queries_unittest.py:
  • Scripts/modules/commands/queues.py:
  • Scripts/modules/mock_bugzillatool.py:
  • Scripts/modules/patchcollection.py:
16:20 Changeset [51463] by abarth@webkit.org

2009-11-28 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] CC webkit-bot-watchers whenever the bots touch bugs
https://bugs.webkit.org/show_bug.cgi?id=31952

The mailing list is open for anyone to subscribe.

  • Scripts/modules/bugzilla.py:
  • Scripts/modules/commands/queues.py:
16:17 Changeset [51462] by abarth@webkit.org

2009-11-28 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] style-queue should report style errors to bugzilla
https://bugs.webkit.org/show_bug.cgi?id=31945

Currently, we're just logging the style errors locally. With this
patch we'll actually log the errors to bugzilla. Note: I plan to run
with the "local-only" logging during development.

  • Scripts/modules/commands/queues.py:
  • Scripts/modules/scm.py:
13:48 Changeset [51461] by eric@webkit.org

2009-11-28 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Eric Seidel.

Apply workaround for the limitation of VirtualFree with MEM_RELEASE to all ports running on Windows
https://bugs.webkit.org/show_bug.cgi?id=31943

  • runtime/MarkStack.h: (JSC::MarkStack::MarkStackArray::shrinkAllocation):
10:27 Changeset [51460] by abarth@webkit.org

2009-11-28 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] style-queue shouldn't reject patches from the commit-queue
https://bugs.webkit.org/show_bug.cgi?id=31944

Currently the style-queue subprocess gets confused and thinks its the
commit-queue. If the patch has an error, it rejects it from the
commit-queue. Instead, we should have style-queue specific logic.
This patch doesn't add that logic, but it gives us a callback we can
use to add that logic.

  • Scripts/modules/buildsteps.py:
  • Scripts/modules/commands/queues.py:
  • Scripts/modules/landingsequence.py:
08:39 Changeset [51459] by abarth@webkit.org

2009-11-28 Adam Barth <abarth@webkit.org>

Reviewed by Dimitri Glazkov.

[Chromium] Sify compose button alerts error
https://bugs.webkit.org/show_bug.cgi?id=31394

Test: http/tests/security/calling-versus-current.html

We're supposed to use the calling context for security checks. In JSC
land, this is the lexicalGlobalObject.

  • bindings/v8/V8Proxy.cpp: (WebCore::V8Proxy::canAccessPrivate):

2009-11-28 Adam Barth <abarth@webkit.org>

Reviewed by Dimitri Glazkov.

[Chromium] Sify compose button alerts error
https://bugs.webkit.org/show_bug.cgi?id=31394

Test that we're using the calling security context in a simple case.

  • http/tests/security/calling-versus-current-expected.txt: Added.
  • http/tests/security/calling-versus-current.html: Added.
07:21 BuildingGtk edited by pierre@alumni.utoronto.ca
(diff)
06:27 Changeset [51458] by eric@webkit.org

2009-11-28 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] SoftwareInputPanelRequest event sent when clicking in newly loaded page

https://bugs.webkit.org/show_bug.cgi?id=31401

Don't set the event unless there is a focused node we can use
for editing afterwards.

  • Api/qwebpage.cpp: (QWebPagePrivate::handleSoftwareInputPanel):
  • tests/qwebpage/tst_qwebpage.cpp: (tst_QWebPage::inputMethods):
02:31 Changeset [51457] by zoltan@webkit.org

2009-11-28 Zoltan Herczeg <zherczeg@inf.u-szeged.hu>

Reviewed by Gavin Barraclough.

https://bugs.webkit.org/show_bug.cgi?id=31930

Seems a typo. We don't need ~270k memory to determine the vptrs.

  • runtime/JSGlobalData.cpp: (JSC::VPtrSet::VPtrSet):
01:09 QtWebKitBackportingFixes edited by hausmann@webkit.org
(diff)
01:07 QtWebKitBackportingFixes edited by hausmann@webkit.org
(diff)

11/27/09:

23:39 Changeset [51456] by yurys@chromium.org

2009-11-27 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Add separate method for pause command to WebDevToolsAgent
and WebDevToolsFrontendClient interfaces so that pause command
can be processed on the IO thread.

https://bugs.webkit.org/show_bug.cgi?id=31934

  • public/WebDevToolsAgent.h:
  • public/WebDevToolsFrontendClient.h: (WebKit::WebDevToolsFrontendClient::sendDebuggerCommandToAgent): (WebKit::WebDevToolsFrontendClient::sendDebuggerPauseScript): (WebKit::WebDevToolsFrontendClient::activateWindow): (WebKit::WebDevToolsFrontendClient::closeWindow): (WebKit::WebDevToolsFrontendClient::dockWindow): (WebKit::WebDevToolsFrontendClient::undockWindow):
22:15 Changeset [51455] by hamaji@chromium.org

2009-11-27 Shinichiro Hamaji <hamaji@chromium.org>

Unreviewed Chromium build fix introduced by r51428.

[Chromium] Ignore line-height CSS property for PushButton
https://bugs.webkit.org/show_bug.cgi?id=31712

  • rendering/RenderThemeChromiumSkia.cpp: (WebCore::RenderThemeChromiumSkia::adjustButtonStyle):
21:33 Changeset [51454] by hamaji@chromium.org

2009-11-27 Shinichiro Hamaji <hamaji@chromium.org>

Unreviewed Chromium test fix by reverting r51413.

[v8] Do not check the thread when accessing DOMDataStore
https://bugs.webkit.org/show_bug.cgi?id=31877

  • bindings/v8/V8DOMMap.cpp: (WebCore::getDOMNodeMap): (WebCore::getDOMObjectMap): (WebCore::getActiveDOMObjectMap): (WebCore::getDOMSVGElementInstanceMap): (WebCore::getDOMSVGObjectWithContextMap):
16:32 Changeset [51453] by dbates@webkit.org

2009-11-27 Daniel Bates <dbates@webkit.org>

Unreviewed, comment fix.

Corrected misspelling of the word "implemented".

  • inspector/front-end/InjectedScriptAccess.js:
16:18 Changeset [51452] by laszlo.1.gombos@nokia.com

2009-11-27 Shinichiro Hamaji <hamaji@chromium.org>

Unreviewed.

Move GOwnPtr* from wtf to wtf/gtk
https://bugs.webkit.org/show_bug.cgi?id=31793

Build fix for chromium after r51423.
Exclude gtk directory from chromium build.

15:46 Changeset [51451] by abarth@webkit.org

Rubber stamped by Eric Seidel.

Rename CommitQueueStatus to QueueStatusServer to allow for more queues.

  • CommitQueueStatus: Removed.
  • CommitQueueStatus/app.yaml: Removed.
  • CommitQueueStatus/filters: Removed.
  • CommitQueueStatus/filters/init.py: Removed.
  • CommitQueueStatus/filters/webkit_extras.py: Removed.
  • CommitQueueStatus/index.html: Removed.
  • CommitQueueStatus/index.yaml: Removed.
  • CommitQueueStatus/queue_status.py: Removed.
  • CommitQueueStatus/stylesheets: Removed.
  • CommitQueueStatus/stylesheets/main.css: Removed.
  • CommitQueueStatus/update_status.html: Removed.
  • QueueStatusServer: Copied from WebKitTools/CommitQueueStatus.
15:39 Changeset [51450] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

style-queue should only process each patch once
https://bugs.webkit.org/show_bug.cgi?id=31939

Actually address reviewer comments!

  • Scripts/bugzilla-tool:
15:36 Changeset [51449] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] style-queue fails to apply a bunch of patches for no reason
https://bugs.webkit.org/show_bug.cgi?id=31942

By passing --non-interactive to check-style, we convince check-style
to pass --force to svn-apply, which lets it apply more patches.

  • Scripts/modules/commands/queues.py:
15:35 Changeset [51448] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Support --status-host in style-queue
https://bugs.webkit.org/show_bug.cgi?id=31941

  • Scripts/modules/commands/queues.py:
  • Scripts/modules/statusbot.py:
15:34 Changeset [51447] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

style-queue should only process each patch once
https://bugs.webkit.org/show_bug.cgi?id=31939

Before processing a patch, the try-queues now ask the web service
whether they have already processed the patch. This is an initial cut
of this functionality. I expect we're make it richer over time.

  • Scripts/bugzilla-tool:
  • Scripts/modules/commands/queues.py:
  • Scripts/modules/patchcollection.py:
15:34 Changeset [51446] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

PatchStatus does not return status
https://bugs.webkit.org/show_bug.cgi?id=31938

We need to parse attachment_id as an int. Otherwise, we can't find
anything in the datastore.

  • CommitQueueStatus/queue_status.py:
15:04 Changeset [51445] by dbates@webkit.org

2009-11-27 Daniel Bates <dbates@webkit.org>

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=31940

Makes the error messages more descriptive when we refuse to load an object/embed or
refuse to load from the document base URL.

  • page/XSSAuditor.cpp: (WebCore::XSSAuditor::canLoadObject): Changed console message to be more descriptive. (WebCore::XSSAuditor::canSetBaseElementURL): Ditto.

2009-11-27 Daniel Bates <dbates@webkit.org>

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=31940

Rebased results because of new error messages.

  • http/tests/security/xssAuditor/base-href-control-char-expected.txt:
  • http/tests/security/xssAuditor/base-href-expected.txt:
  • http/tests/security/xssAuditor/base-href-null-char-expected.txt:
  • http/tests/security/xssAuditor/base-href-scheme-relative-expected.txt:
  • http/tests/security/xssAuditor/embed-tag-control-char-expected.txt:
  • http/tests/security/xssAuditor/embed-tag-expected.txt:
  • http/tests/security/xssAuditor/embed-tag-javascript-url-expected.txt:
  • http/tests/security/xssAuditor/embed-tag-null-char-expected.txt:
  • http/tests/security/xssAuditor/object-embed-tag-control-char-expected.txt:
  • http/tests/security/xssAuditor/object-embed-tag-expected.txt:
  • http/tests/security/xssAuditor/object-embed-tag-null-char-expected.txt:
  • http/tests/security/xssAuditor/object-src-inject-expected.txt:
  • http/tests/security/xssAuditor/object-tag-expected.txt:
  • http/tests/security/xssAuditor/object-tag-javascript-url-expected.txt:
10:49 Changeset [51444] by eric@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Make commit-queue status not throw an exception where there is not
status
https://bugs.webkit.org/show_bug.cgi?id=31936

We need to actually fetch the results of the query to see what's in the
datastore.

  • CommitQueueStatus/queue_status.py:
10:09 Changeset [51443] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] unit test ApplyPatches and ApplyAttachment
https://bugs.webkit.org/show_bug.cgi?id=31935

  • Scripts/modules/commands/download.py:
  • Scripts/modules/commands/download_unittest.py:
09:37 Changeset [51442] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Unit test download commands
https://bugs.webkit.org/show_bug.cgi?id=31923

Adds download_unittest and fixes a bug found while testing.

  • Scripts/modules/commands/commandtest.py:
  • Scripts/modules/commands/download.py:

Fixed a bug where we'd throw an error because [].append returns
None.

  • Scripts/modules/commands/download_unittest.py: Added.
  • Scripts/modules/mock_bugzillatool.py:
  • Scripts/run-webkit-unittests:
08:21 Changeset [51441] by eric@webkit.org

2009-11-27 Kenneth Russell <kbr@google.com>

Reviewed by Eric Seidel.

REGRESSION(51418): fast/canvas/webgl/gl-object-get-calls.html is failing on the Leopard commit bot
https://bugs.webkit.org/show_bug.cgi?id=31919

Temporarily disabled portion of test which is failing on some
machines but not others. Root cause will be investigated under
https://bugs.webkit.org/show_bug.cgi?id=31843 .

  • fast/canvas/webgl/gl-object-get-calls-expected.txt:
  • fast/canvas/webgl/script-tests/gl-object-get-calls.js:
05:07 Changeset [51440] by yurys@chromium.org

2009-11-27 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

This is a WebCore part of the fix that allows to view plugin
resources loaded by plugins.

https://bugs.webkit.org/show_bug.cgi?id=31832

  • inspector/InspectorController.cpp: (WebCore::InspectorController::willSendRequest): (WebCore::InspectorController::didReceiveResponse): (WebCore::InspectorController::didReceiveContentLength): (WebCore::InspectorController::didFinishLoading): (WebCore::InspectorController::didFailLoading):
  • inspector/InspectorController.h:
  • inspector/InspectorResource.cpp: (WebCore::InspectorResource::addLength):
  • loader/ResourceLoadNotifier.cpp: (WebCore::ResourceLoadNotifier::didFailToLoad): (WebCore::ResourceLoadNotifier::dispatchWillSendRequest): (WebCore::ResourceLoadNotifier::dispatchDidReceiveResponse): (WebCore::ResourceLoadNotifier::dispatchDidReceiveContentLength): (WebCore::ResourceLoadNotifier::dispatchDidFinishLoading):

2009-11-27 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

WebCore part of the fix that allows to view plugin
resources loaded by plugins. Methods that for resource
loading notifications are exposed through WebDevToolsAgent
interface to the glue code to allow Chromium's plugin
implementation to notify InspectorController about resource
loading in plugins.

https://bugs.webkit.org/show_bug.cgi?id=31832

  • public/WebDevToolsAgent.h:
  • public/WebView.h:
  • src/WebViewImpl.cpp: (WebKit::WebViewImpl::createUniqueIdentifierForRequest):
  • src/WebViewImpl.h:
03:12 Changeset [51439] by yurys@chromium.org

2009-11-27 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Instead of generating negative identifiers for cached resources in
InspectorController and extending identifier type from 'unsigned long' to 'long
long' reuse progress tracker from the inspected page to generate those
identifiers. It guarantees that InspectorResources have unique ids since
all of them are generated by that progress tracker.

Added a couple new overloaded methods to Script* objects that accept
long and unsigned long arguments. These types of argumens have already
been passed as long long.

https://bugs.webkit.org/show_bug.cgi?id=31921

  • bindings/js/ScriptFunctionCall.cpp: (WebCore::ScriptFunctionCall::appendArgument):
  • bindings/js/ScriptFunctionCall.h:
  • bindings/js/ScriptObject.cpp: (WebCore::ScriptObject::set):
  • bindings/js/ScriptObject.h:
  • bindings/v8/ScriptFunctionCall.cpp: (WebCore::ScriptFunctionCall::appendArgument):
  • bindings/v8/ScriptFunctionCall.h:
  • bindings/v8/ScriptObject.cpp: (WebCore::ScriptObject::set):
  • bindings/v8/ScriptObject.h:
  • inspector/InspectorBackend.cpp: (WebCore::InspectorBackend::dispatchOnInjectedScript):
  • inspector/InspectorController.cpp: (WebCore::InspectorController::InspectorController): (WebCore::InspectorController::getTrackedResource): (WebCore::InspectorController::didLoadResourceFromMemoryCache): Use inspected page's ProgressTracker to generate unique identifiers for cached resources in InspectorController.
  • inspector/InspectorController.h:
  • inspector/InspectorDOMAgent.cpp: (WebCore::InspectorDOMAgent::buildObjectForNode): (WebCore::InspectorDOMAgent::buildObjectForEventListener):
  • inspector/InspectorFrontend.cpp: (WebCore::InspectorFrontend::addResource): (WebCore::InspectorFrontend::updateResource): (WebCore::InspectorFrontend::removeResource): (WebCore::InspectorFrontend::updateFocusedNode):
  • inspector/InspectorFrontend.h:
  • inspector/InspectorResource.cpp: (WebCore::InspectorResource::InspectorResource): (WebCore::InspectorResource::createCached):
  • inspector/InspectorResource.h: Change InspectorResource identifier type from 'long long' to 'unsigned long'. (WebCore::InspectorResource::create): (WebCore::InspectorResource::identifier):
  • inspector/TimelineRecordFactory.cpp: (WebCore::TimelineRecordFactory::createResourceSendRequestRecord): (WebCore::TimelineRecordFactory::createResourceReceiveResponseRecord): (WebCore::TimelineRecordFactory::createResourceFinishRecord):
00:45 Changeset [51438] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Unreviewed "build" fix found while writing unit tests.

  • Scripts/modules/commands/download.py:
00:04 Changeset [51437] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Kill WebKitLandingScripts
https://bugs.webkit.org/show_bug.cgi?id=31904

Step 6: Kill the rest.

  • Scripts/modules/buildsteps.py:
  • Scripts/modules/commands/download.py:
  • Scripts/modules/commands/queries.py:
  • Scripts/modules/commands/upload.py:
  • Scripts/modules/landingsequence.py:
  • Scripts/modules/scm.py:
  • Scripts/modules/webkitlandingscripts.py: Removed.
00:02 Changeset [51436] by abarth@webkit.org

2009-11-27 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Unit test upload commands
https://bugs.webkit.org/show_bug.cgi?id=31903

Adds unit tests for all but two of the upload commands. The two
remaining ones are more difficult. I'll return to them later. The
goal of these tests is just to run the commands. We can test more
detailed behavior later.

  • Scripts/modules/commands/commandtest.py:
  • Scripts/modules/commands/upload.py:
  • Scripts/modules/commands/upload_unittest.py:
  • Scripts/modules/mock.py: Added.
  • Scripts/modules/mock_bugzillatool.py:

11/26/09:

23:56 Changeset [51435] by abarth@webkit.org

2009-11-26 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Kill WebKitLandingScripts
https://bugs.webkit.org/show_bug.cgi?id=31904

Step 5: Kill run_and_throw_if_fail.

  • Scripts/modules/buildsteps.py:
  • Scripts/modules/processutils.py: Added.
  • Scripts/modules/webkitlandingscripts.py:
23:56 Changeset [51434] by abarth@webkit.org

2009-11-26 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Kill WebKitLandingScripts
https://bugs.webkit.org/show_bug.cgi?id=31904

Step 4: Kill run_webkit_script.

  • Scripts/modules/buildsteps.py:
  • Scripts/modules/commands/download.py:
  • Scripts/modules/webkitlandingscripts.py:
23:55 Changeset [51433] by abarth@webkit.org

2009-11-26 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Kill WebKitLandingScripts
https://bugs.webkit.org/show_bug.cgi?id=31904

Step 3: Kill build_webkit.

  • Scripts/modules/buildsteps.py:
  • Scripts/modules/landingsequence.py:
  • Scripts/modules/webkitlandingscripts.py:
23:55 Changeset [51432] by abarth@webkit.org

2009-11-26 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Kill WebKitLandingScripts
https://bugs.webkit.org/show_bug.cgi?id=31904

Step 2: Kill ensure_builders_are_green.

  • Scripts/modules/buildsteps.py:
  • Scripts/modules/commands/download.py:
  • Scripts/modules/landingsequence.py:
  • Scripts/modules/webkitlandingscripts.py:
23:54 Changeset [51431] by abarth@webkit.org

2009-11-26 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Kill WebKitLandingScripts
https://bugs.webkit.org/show_bug.cgi?id=31904

Step 1: Kill prepare_clean_working_directory and run_webkit_tests.

  • Scripts/bugzilla-tool:
  • Scripts/modules/buildsteps.py: Added.
  • Scripts/modules/commands/download.py:
  • Scripts/modules/landingsequence.py:
  • Scripts/modules/webkitlandingscripts.py:
23:51 Changeset [51430] by hamaji@chromium.org

2009-11-26 Shinichiro Hamaji <hamaji@chromium.org>

Unreviewed red bot fix.

Skip tests introduced by the previous change as it seems to be failing.

  • platform/qt/Skipped:
  • platform/win/Skipped:
22:11 Changeset [51429] by hamaji@chromium.org

2009-11-26 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Dan Bernstein.

Assertion failure in RenderBlock::positionForPointWithInlineChildren when running fast/inline/relative-positioned-overflow.html
https://bugs.webkit.org/show_bug.cgi?id=29966

A test case where a user drags the mouse from an empty inline
element to above texts.

  • editing/selection/last-empty-inline-expected.txt: Added.
  • editing/selection/last-empty-inline.html: Added.
  • platform/win/Skipped:

2009-11-26 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Dan Bernstein.

Assertion failure in RenderBlock::positionForPointWithInlineChildren when running fast/inline/relative-positioned-overflow.html
https://bugs.webkit.org/show_bug.cgi?id=29966

When an empty inline element is clicked, the root inline box has
no leaf children. Use the renderer of a normal child instead.

This change resolves Windows port's assertion failure in a layout test.
Also, this fixes the behavior when a user drags the mouse from an
empty inline element to above texts.

Test: editing/selection/last-empty-inline.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::positionForPointWithInlineChildren):
21:59 Changeset [51428] by eric@webkit.org

2009-11-26 Kent Tamura <tkent@chromium.org>

Reviewed by Eric Seidel.

[Chromium] Ignore line-height CSS property specified to push buttons on
Windows and Linux.
https://bugs.webkit.org/show_bug.cgi?id=31712

LayoutTests/fast/forms/control-restrict-line-height.html checks that the
following controls should ignore line-height CSS property.

  • <select>
  • <input type=button>
  • <input type=search>

This change addresses the <input type=button> issue with Chromium/Windows
and Chromium/Linux.

  • rendering/RenderThemeChromiumSkia.cpp: Implement adjustButtonStyle() to ignore line-height.
  • rendering/RenderThemeChromiumSkia.h: Declare adjustButtonStyle().
21:53 Changeset [51427] by eric@webkit.org

2009-11-26 Kinuko Yasuda <kinuko@chromium.com>

Reviewed by Eric Seidel.

Rewrite the test fast/parser/comment-in-textarea to use dumpAsText
so that the expected results can be independent of platform differences.
https://bugs.webkit.org/show_bug.cgi?id=31580

  • fast/parser/comment-in-textarea-expected.txt: Added.
  • fast/parser/comment-in-textarea.html:
  • fast/parser/script-tests/comment-in-textarea.js: Added.
  • platform/mac/fast/parser/comment-in-textarea-expected.checksum: Removed.
  • platform/mac/fast/parser/comment-in-textarea-expected.png: Removed.
  • platform/mac/fast/parser/comment-in-textarea-expected.txt: Removed.
  • platform/win/fast/parser/comment-in-textarea-expected.txt: Removed.
21:46 Changeset [51426] by eric@webkit.org

2009-11-26 Kinuko Yasuda <kinuko@chromium.com>

Reviewed by Eric Seidel.

Remove the special charset meta tag in the clipboard so that
copy-and-paste with interchange breaklines/spaces works correctly
within WebKit.

No new tests. Layout tests that involve copy-and-paste with
interchange breaklines (like editing/pasteboard/paste-line-endings-00?)
should pass on Mac/Chromium with this fix.

  • platform/chromium/ClipboardChromium.cpp:
  • platform/chromium/PasteboardChromium.cpp: (WebCore::Pasteboard::documentFragment):
21:33 Changeset [51425] by eric@webkit.org

2009-11-26 İsmail Dönmez <ismail@namtrac.org>

Reviewed by Eric Seidel.

Compile with QT_NO_WHEELEVENT defined.

  • platform/qt/WheelEventQt.cpp: (WebCore::PlatformWheelEvent::applyDelta):
21:20 Changeset [51424] by oliver@apple.com

Incorrect behaviour of jneq_null in the interpreter
https://bugs.webkit.org/show_bug.cgi?id=31901

Reviewed by Gavin Barraclough.

Correct the logic of jneq_null. This is already covered by existing tests.

20:06 Changeset [51423] by laszlo.1.gombos@nokia.com

2009-11-26 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Oliver Hunt.

Move GOwnPtr* from wtf to wtf/gtk
https://bugs.webkit.org/show_bug.cgi?id=31793

  • GNUmakefile.am: Change the path for GOwnPtr.*.
  • JavaScriptCore.gyp/JavaScriptCore.gyp: Remove GOwnPtr.cpp from the exclude list.
  • JavaScriptCore.gypi: Change the path for GOwnPtr.*.
  • wscript: Remove GOwnPtr.cpp from the exclude list.
  • wtf/GOwnPtr.cpp: Removed.
  • wtf/GOwnPtr.h: Removed.
  • wtf/Threading.h: Change the path for GOwnPtr.h.
  • wtf/gtk/GOwnPtr.cpp: Copied from JavaScriptCore/wtf/GOwnPtr.cpp.
  • wtf/gtk/GOwnPtr.h: Copied from JavaScriptCore/wtf/GOwnPtr.h.
  • wtf/unicode/glib/UnicodeGLib.h: Change the path for GOwnPtr.h.

2009-11-26 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Oliver Hunt.

Move GOwnPtr* from wtf to wtf/gtk
https://bugs.webkit.org/show_bug.cgi?id=31793

  • webkit/webkitwebview.cpp: Change the path for GOwnPtr.h.

2009-11-26 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Oliver Hunt.

Move GOwnPtr* from wtf to wtf/gtk
https://bugs.webkit.org/show_bug.cgi?id=31793

No new tests as there is no functional change.

  • platform/graphics/gtk/MediaPlayerPrivateGStreamer.cpp: Change the path for GOwnPtr.h.
  • platform/text/TextEncoding.cpp: Ditto.
  • platform/text/gtk/TextCodecGtk.cpp: Ditto.

2009-11-26 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Oliver Hunt.

Move GOwnPtr* from wtf to wtf/gtk
https://bugs.webkit.org/show_bug.cgi?id=31793

13:22 Changeset [51422] by yurys@chromium.org

2009-11-26 Yury Semikhatsky <yurys@chromium.org>

Not reviewed. Build fix: revert r51421.

  • inspector/InspectorController.cpp: (WebCore::InspectorController::willSendRequest): (WebCore::InspectorController::didReceiveResponse): (WebCore::InspectorController::didReceiveContentLength): (WebCore::InspectorController::didFinishLoading): (WebCore::InspectorController::didFailLoading):
  • inspector/InspectorController.h:
  • inspector/InspectorResource.cpp: (WebCore::InspectorResource::addLength):
  • loader/ResourceLoadNotifier.cpp: (WebCore::ResourceLoadNotifier::didFailToLoad): (WebCore::ResourceLoadNotifier::dispatchWillSendRequest): (WebCore::ResourceLoadNotifier::dispatchDidReceiveResponse): (WebCore::ResourceLoadNotifier::dispatchDidReceiveContentLength): (WebCore::ResourceLoadNotifier::dispatchDidFinishLoading):

2009-11-26 Yury Semikhatsky <yurys@chromium.org>

Not reviewed. Build fix: revert r51421.

  • public/WebDevToolsAgent.h:
  • public/WebView.h:
  • src/WebViewImpl.cpp:
  • src/WebViewImpl.h:
12:59 Changeset [51421] by yurys@chromium.org

2009-11-26 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

This is a WebCore part of the fix that allows to view plugin
resources loaded by plugins.

https://bugs.webkit.org/show_bug.cgi?id=31832

  • inspector/InspectorController.cpp: (WebCore::InspectorController::willSendRequest): (WebCore::InspectorController::didReceiveResponse): (WebCore::InspectorController::didReceiveContentLength): (WebCore::InspectorController::didFinishLoading): (WebCore::InspectorController::didFailLoading):
  • inspector/InspectorController.h: Remove unused DocumentLoader parameters from inspector methods.
  • inspector/InspectorResource.cpp: (WebCore::InspectorResource::addLength): Update loading end time when new data are received.
  • loader/ResourceLoadNotifier.cpp: (WebCore::ResourceLoadNotifier::didFailToLoad): Notify InspectorController about the failure. (WebCore::ResourceLoadNotifier::dispatchWillSendRequest): (WebCore::ResourceLoadNotifier::dispatchDidReceiveResponse): (WebCore::ResourceLoadNotifier::dispatchDidReceiveContentLength): (WebCore::ResourceLoadNotifier::dispatchDidFinishLoading):

2009-11-26 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

WebCore part of the fix that allows to view plugin
resources loaded by plugins. Methods that for resource
loading notifications are exposed through WebDevToolsAgent
interface to the glue code to allow Chromium's plugin
implementation to notify InspectorController about resource
loading in plugins.

https://bugs.webkit.org/show_bug.cgi?id=31832

  • public/WebDevToolsAgent.h:
  • public/WebView.h:
  • src/WebViewImpl.cpp: (WebKit::WebViewImpl::createUniqueIdentifierForRequest):
  • src/WebViewImpl.h:
12:27 Changeset [51420] by eric@webkit.org

2009-11-26 Daniel Bates <dbates@webkit.org>

Reviewed by Pavel Feldman.

https://bugs.webkit.org/show_bug.cgi?id=21554

Tests that the tooltip text for an image is its dimensions.

  • inspector/elements-img-tooltip-expected.txt: Added.
  • inspector/elements-img-tooltip.html: Added.

2009-11-26 Daniel Bates <dbates@webkit.org>

Reviewed by Pavel Feldman.

https://bugs.webkit.org/show_bug.cgi?id=21554

Implements support for hovering over <img> src to display the height and width of that image
in a tooltip. Displays both the displayable and natural dimensions of the image.

Test: inspector/elements-img-tooltip.html

  • inspector/front-end/ElementsTreeOutline.js: (WebInspector.ElementsTreeElement.prototype.createTooltipForImageNode): Added. (WebInspector.ElementsTreeElement.prototype._updateTitle.callback): (WebInspector.ElementsTreeElement.prototype._updateTitle): (WebInspector.ElementsTreeElement.prototype._nodeTitleInfo):
  • inspector/front-end/ObjectProxy.js: (WebInspector.ObjectProxy.getPropertiesAsync): Added.
  • inspector/front-end/inspector.js: (WebInspector.linkifyURLAsNode): Added tooltipText argument. (WebInspector.linkifyURL): Ditto.
10:06 Changeset [51419] by kevino@webkit.org

wx build fix after drawPattern API change.

08:50 Changeset [51418] by cmarrin@apple.com

Missed checking in these files from https://bugs.webkit.org/show_bug.cgi?id=30091

08:50 Changeset [51417] by zoltan@webkit.org

2009-11-26 Zoltan Horvath <zoltan@webkit.org>

Reviewed by Adam Barth.

Remove directory prefixes from linux commands

SCM unittests can cause errors on some systems if we use absolute reference to
the commands, so directory prefixes have been removed.

  • Scripts/modules/scm_unittest.py:
07:59 Changeset [51416] by eric@webkit.org

2009-11-26 Anton Muhin <antonm@chromium.org>

Reviewed by Adam Barth.

Use an internal field instead of hidden property to speedup lookup
of entered isolated world.

Plus some inlinings.
https://bugs.webkit.org/show_bug.cgi?id=31884

Covered by layout tests + manual running of some benchmarks as
content scripts.

  • bindings/v8/V8DOMWrapper.cpp:
  • bindings/v8/V8DOMWrapper.h: (WebCore::V8DOMWrapper::setDOMWrapper):
  • bindings/v8/V8HiddenPropertyName.h:
  • bindings/v8/V8IsolatedWorld.cpp: (WebCore::V8IsolatedWorld::V8IsolatedWorld):
  • bindings/v8/V8IsolatedWorld.h: (WebCore::V8IsolatedWorld::getEntered): (WebCore::V8IsolatedWorld::getGlobalObject):
  • bindings/v8/custom/V8CustomBinding.h:
07:06 Changeset [51415] by hausmann@webkit.org

Rubber-stamped by Holger Freyther.

Removed unused ICO image plugin handler.

  • Plugins/ICOHandler.cpp: Removed.
  • Plugins/ICOHandler.h: Removed.
  • Plugins/Plugins.pro: Removed.
06:34 Changeset [51414] by eric@webkit.org

2009-11-26 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Corrects build break on Windows.

Rename platform/text/qt/TextBoundaries.cpp to TextBoundariesQt.cpp since
platform/text/TextBoundaries.cpp was compiled instead when compiling with nmake.

  • WebCore.pro:
  • platform/text/qt/TextBoundariesQt.cpp: Renamed from WebCore/platform/text/qt/TextBoundaries.cpp. (WebCore::findNextWordFromIndex): (WebCore::findWordBoundary):
05:20 Changeset [51413] by eric@webkit.org

2009-11-26 Anton Muhin <antonm@chromium.org>

Reviewed by Adam Barth.

Do not check if the thread is main or not when accessing DOMDataStore as currently in Chromium WebKit is used in main thread only.
https://bugs.webkit.org/show_bug.cgi?id=31877

Covered by layout tests and buildbots.

  • bindings/v8/V8DOMMap.cpp: (WebCore::getDOMNodeMap): (WebCore::getDOMObjectMap): (WebCore::getActiveDOMObjectMap): (WebCore::getDOMSVGElementInstanceMap): (WebCore::getDOMSVGObjectWithContextMap):
05:04 Changeset [51412] by girish@forwardbias.in

[Qt] Mac Plugins : Pass mouse position relative to the fake window

Reviewed by Holger Freyther.

When using off-screen rendering, we need to pass mouse events relative
to the fake window instead of the global position.

https://bugs.webkit.org/show_bug.cgi?id=31794

04:44 QtWebKitMirrorGuide edited by zecke@selfish.org
Mention the Math.random() problem with the test content. (diff)
04:33 QtWebKitPerformanceToolBackLog: edited by zecke@selfish.org
(diff)
04:31 QtWebKitPerformanceToolBackLog: edited by zecke@selfish.org
(diff)
04:26 QtWebKitPerformanceToolBackLog: edited by zecke@selfish.org
Fix the marking of completed items. (diff)
04:23 QtBackLog edited by zecke@selfish.org
Put the commits and date into the right column (diff)
04:21 QtBackLog edited by zecke@selfish.org
Mention the network change, and the issue that QtWebKit is using less … (diff)
04:18 QtBackLog edited by zecke@selfish.org
Complete the QtPlugin work (diff)
03:16 Changeset [51411] by zecke@webkit.org

[Qt] Use QNetworkReply::rawHeaderPairs

https://bugs.webkit.org/show_bug.cgi?id=31826

The QNetworkReply is internally storing the HTTP headers
as a list of pairs. Currently we have to ask the QNetworkReply
to put all header names into a QStringList. Afterwards we will
iterate over this QStringList and ask the QNetworkReply to
give us the value for this header name. The current Qt implementation
is doing a linear to find the header value.

Use a new API to directly access the list of pairs and push
this into WebCore. This avoids doing some allocations and doing
linear searches from within a loop.

  • platform/network/qt/QNetworkReplyHandler.cpp:

(WebCore::QNetworkReplyHandler::sendResponseIfNeeded):

02:34 Changeset [51410] by zecke@webkit.org

[Qt] Add automatic test case for plugins that use QWidget

https://bugs.webkit.org/show_bug.cgi?id=31203

Change the RenderTreeAsText to look into the WebCore::Widget
and print the state of the platformWidget. Change the DRT to
use QtUiTools to be able to initiate QProgressBar and such
as a plugin. Add the test case, add the result, update the
other results due the new test.

WebCore:

  • rendering/RenderTreeAsText.cpp:

(WebCore::operator<<): Add special case for RenderPart

WebKitTools:

  • DumpRenderTree/qt/DumpRenderTree.cpp:

(WebCore::WebPage::createPlugin):

  • DumpRenderTree/qt/DumpRenderTree.h:
  • DumpRenderTree/qt/DumpRenderTree.pro:

LayoutTest:

  • platform/gtk/Skipped:
  • platform/mac/Skipped:
  • platform/qt/plugins/netscape-dom-access-expected.txt: Added
  • platform/qt/plugins/qt-qwidget-plugin-expected.txt: Added
  • platform/win/Skipped:
  • plugins/qt-qwidget-plugin.html: Added.
02:31 Changeset [51409] by zecke@webkit.org

[Qt] Call Widget::setSelfVisible from hide/show

https://bugs.webkit.org/show_bug.cgi?id=31203

Call Widget::setSelfVisible from Widget::show and
Widget::hide and use isParentVisible to decide
if the widget should be shown. This way client
code can rely on isVisible.

Change PluginViewQt::show, PluginViewQt::hide to
call the base class as it is doing the right thing
now. Add an assert verify that platfomWidget and
platformPluginWidget are the same.

  • manual-tests/qt/qtplugin.html: Modify manual test
  • platform/qt/WidgetQt.cpp:

(WebCore::Widget::show):
(WebCore::Widget::hide):

[Qt] Do not show the QWidget when the WebCore::Widget is hidden

https://bugs.webkit.org/show_bug.cgi?id=31203

The clipping code was making a QWidget visible even if the
WebCore::Widget was hidden. Fix the bug by calling setVisible
only if the WebCore::Widget Widget::isSelfVisible.

  • WebCoreSupport/FrameLoaderClientQt.cpp:

(WebCore::QtPluginWidget::show): Override WebCore::Widget::show to call handleVisibility
(WebCore::QtPluginWidget::handleVisibility): New method to call setVisible when we are visible
(FrameLoaderClientQt::createPlugin): Hide the QWidget by default

02:31 Changeset [51408] by zecke@webkit.org

Fix compilation of REQUEST_DEBUG debug code

https://bugs.webkit.org/show_bug.cgi?id=31850

In r47907 the single parameter KURL constructor to parse
from a WebCore::String was replaced with a two parameter
constructor. I think in this debug case parsing the urls
again is no problem and I have changed the code to use the
two parameter version.

  • loader/loader.cpp:

(WebCore::Loader::Host::didFinishLoading):

00:43 Changeset [51407] by eric@webkit.org

2009-11-26 Søren Gjesse <sgjesse@chromium.org>

Reviewed by Pavel Feldman.

[V8] Avoid using JavaScript objects as context data
https://bugs.webkit.org/show_bug.cgi?id=31873

Change the context "data" from a JavaScript object holding the two properties type and value to
a string holding type and value separated by a comma.

  • bindings/v8/V8Proxy.cpp: (WebCore::V8Proxy::setInjectedScriptContextDebugId): (WebCore::V8Proxy::setContextDebugId): (WebCore::V8Proxy::contextDebugId):

11/25/09:

22:16 Changeset [51406] by eric@webkit.org

2009-11-25 Yuzo Fujishima <yuzo@google.com>

Reviewed by Eric Seidel.

Update pywebsocket to 0.4.2

Update pywebsocket to 0.4.2
https://bugs.webkit.org/show_bug.cgi?id=31861

  • pywebsocket/example/echo_client.py:
  • pywebsocket/example/echo_wsh.py:
  • pywebsocket/mod_pywebsocket/init.py:
  • pywebsocket/mod_pywebsocket/dispatch.py:
  • pywebsocket/mod_pywebsocket/msgutil.py:
  • pywebsocket/mod_pywebsocket/standalone.py:
  • pywebsocket/setup.py:
  • pywebsocket/test/test_dispatch.py:
  • pywebsocket/test/test_msgutil.py:
19:02 Changeset [51405] by abarth@webkit.org

2009-11-25 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Add unit test for mark-fixed
https://bugs.webkit.org/show_bug.cgi?id=31896

  • Scripts/modules/commands/commandtest.py: Added.
  • Scripts/modules/commands/queries_unittest.py:
  • Scripts/modules/commands/upload_unittest.py: Added.
  • Scripts/modules/mock_bugzillatool.py:
  • Scripts/run-webkit-unittests:
18:25 Changeset [51404] by abarth@webkit.org

2009-11-25 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

bugzilla-tool should have a mark-fixed command
https://bugs.webkit.org/show_bug.cgi?id=31853

Pretty simple stuff.

  • Scripts/modules/commands/upload.py:
18:22 Changeset [51403] by eric@webkit.org

2009-11-25 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

'bugzilla-tool help' should only show common commands like how 'git help' does
https://bugs.webkit.org/show_bug.cgi?id=31772

I also took this opportunity to make 'help' a real Command.
Making 'help' a real command required adding Command.tool (which we've wanted to do for a while).

  • Scripts/bugzilla-tool:
    • change should_show_command_help to should_show_in_main_help
  • Scripts/modules/commands/download.py:
    • Mark commands as being shown in main help or not.
    • show_in_main_help = False is not required (default is false), but it seemed to make the commands more self-documenting.
  • Scripts/modules/commands/queries.py: ditto
  • Scripts/modules/commands/queues.py: ditto
  • Scripts/modules/commands/upload.py: ditto
  • Scripts/modules/multicommandtool.py:
    • Make Command hold a pointer to tool in self.tool. Most Command subclasses do not take advantage of this yet, but it was required for HelpCommand to be able to reach the tool from _help_epilog().
    • Move MultiCommandTool._standalone_help_for_command to Command.standalone_help
    • Move MultiCommandTool._help_epilog to Command._help_epilog
    • Move "help" logic into HelpCommand.execute()
    • Change should_show_command_help to should_show_in_main_help and add a default implementation.
  • Scripts/modules/multicommandtool_unittest.py:
    • Test hiding of Commands in --help, and that all commands are shown in 'help --all-commands'
16:57 Changeset [51402] by dglazkov@chromium.org

WebCore:

2009-11-25 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by David Levin.

[Chromium] Implement canSetValueAttribute in the API, the clean-up part.
https://bugs.webkit.org/show_bug.cgi?id=31894

  • accessibility/chromium/AccessibilityObjectWrapper.h: Added RefCounted decl. (WebCore::AccessibilityObjectWrapper::AccessibilityObjectWrapper): Removed mis-refcountingness.

WebKit/chromium:

2009-11-25 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by David Levin.

[Chromium] Implement canSetValueAttribute in the API.
https://bugs.webkit.org/show_bug.cgi?id=31894

  • public/WebAccessibilityObject.h: Added decl.
  • src/WebAccessibilityCacheImpl.h: Cleaned up mis-include.
  • src/WebAccessibilityObject.cpp: (WebKit::WebAccessibilityObject::canSetValueAttribute): Added impl.
16:51 Changeset [51401] by ossy@webkit.org

[Qt] Audio related layout tests are failing on the buildbot
https://bugs.webkit.org/show_bug.cgi?id=27910

  • platform/qt/Skipped: fast/js/global-constructors.html added.
16:44 Changeset [51400] by eric@webkit.org

2009-11-25 Kenneth Russell <kbr@google.com>

Reviewed by Oliver Hunt.

Off-by-one error in index validation for drawElements and drawArrays
https://bugs.webkit.org/show_bug.cgi?id=31891

Fixed computation of number of elements for bound array objects.

Test: fast/canvas/webgl/index-validation.html

  • fast/canvas/webgl/index-validation-expected.txt: Added.
  • fast/canvas/webgl/index-validation.html: Added.
  • fast/canvas/webgl/script-tests/index-validation.js: Added.

2009-11-25 Kenneth Russell <kbr@google.com>

Reviewed by Oliver Hunt.

Off-by-one error in index validation for drawElements and drawArrays
https://bugs.webkit.org/show_bug.cgi?id=31891

Fixed computation of number of elements for bound array objects.

Test: fast/canvas/webgl/index-validation.html

  • html/canvas/WebGLRenderingContext.cpp: (WebCore::WebGLRenderingContext::vertexAttribPointer):
16:41 Changeset [51399] by bweinstein@apple.com

Patch by Mark Rowe.

Reviewed by Dan Bernstein.

The buildbots are failing on Windows because when they were upgraded
to 4.0.4, Apple Application Support was not in their path. Add it to
the path to fix the buildots.

  • Scripts/webkitdirs.pm:
16:29 Changeset [51398] by ossy@webkit.org

[Qt] Remove a bunch of now passing fast tests from skiplist.
And add a crasher/crashed pair into skiplist.

  • platform/qt/Skipped:
15:44 Changeset [51397] by dimich@chromium.org

Update SharedScript to use eventNames() instead of EventNames()
https://bugs.webkit.org/show_bug.cgi?id=31890

Reviewed by David Levin.

  • SharedScript/WebKitSharedScript.cpp:

(WebCore::LoadEventTask::performTask):

15:29 Changeset [51396] by ossy@webkit.org

[Qt] Put tests into skiplist because of missing layoutTestController feature.

  • platform/qt/Skipped:
  • http/tests/cache/subresource-failover-to-network.html added.
  • fast/xmlhttprequest/xmlhttprequest-nonexistent-file.html added.
  • "3D_RENDERING and ACCELERATED_COMPOSITING" block moved to good place.
14:26 Changeset [51395] by ossy@webkit.org

run-webkit-tests doesn't accept directories/files with --skipped=only parameter
https://bugs.webkit.org/show_bug.cgi?id=31799

Reviewed by David Kilzer.

  • Scripts/run-webkit-tests: Fixed.
13:09 Changeset [51394] by darin@chromium.org

2009-11-25 Darin Fisher <darin@chromium.org>

Reviewed by Adam Barth.

[Chromium] Unable to repost form data by hitting reload
https://bugs.webkit.org/show_bug.cgi?id=31882

  • src/WebFrameImpl.cpp: (WebKit::WebFrameImpl::loadData):
13:08 Changeset [51393] by eric.carlson@apple.com

2009-11-25 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

<rdar://problem/7409331> Windows: Support closed caption in <video> element

Enable closed captions in QuickTime/Windows media engine.

  • platform/graphics/win/MediaPlayerPrivateQuickTimeWin.cpp: (WebCore::MediaPlayerPrivate::hasClosedCaptions): (WebCore::MediaPlayerPrivate::setClosedCaptionsVisible):

New, all through to m_qtMovie.

  • platform/graphics/win/MediaPlayerPrivateQuickTimeWin.h:

Make all but the destructor private since MediaPlayer call through the media
engine interface.

  • platform/graphics/win/QTMovieWin.cpp: (QTMovieWin::disableUnsupportedTracks):

Use handy new constants for QuickTime track types.

(QTMovieWin::hasClosedCaptions):
(QTMovieWin::setClosedCaptionsVisible):

New, closed caption support.

  • platform/graphics/win/QTMovieWin.h:
  • rendering/RenderMediaControls.cpp: (WebCore::RenderMediaControls::paintMediaControlsPart):

Deal with closed caption buttons.

  • rendering/RenderThemeWin.cpp: (WebCore::RenderThemeWin::shouldRenderMediaControlPart):

New, don't ask the media engine if it has closed captions unless the Safari theme will
be able to render the button.

(WebCore::RenderThemeWin::paintMediaToggleClosedCaptionsButton):

New.

  • rendering/RenderThemeWin.h:

2009-11-25 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

<rdar://problem/7409331> Windows: Support closed caption in <video> element

  • platform/win/Skipped:

Enable the closed caption test.

12:04 Changeset [51392] by eric@webkit.org

2009-11-25 Drew Wilson <atwilson@chromium.org>

Reviewed by David Levin.

MessagePorts always look remotely entangled even when closed.
https://bugs.webkit.org/show_bug.cgi?id=31698

Tests: Existing tests suffice, Chromium soak test passes now.

  • bindings/v8/custom/V8CustomBinding.h: Removed kMessagePortEntangledPortIndex which is no longer used.
  • bindings/v8/V8GCController.cpp: (WebCore::GCPrologueVisitor::visitDOMWrapper): Simplified GC code to reflect the Chromium MessagePort implementation (locallyEntangledPort() always returns false). (WebCore::GCEpilogueVisitor::visitDOMWrapper): Cleaned up epilogue code to handle the case where the port gets closed in mid-GC (due to the parent context being freed).
  • dom/MessagePort.cpp: (WebCore::MessagePort::MessagePort): (WebCore::MessagePort::close): Now sets the closed flag. (WebCore::MessagePort::disentanglePorts): Updated to use new isCloned() API instead of relying on isEntangled(), which was incorrect.
  • dom/MessagePort.h: Added a m_closed flag and updated isEntangled() to check it. (WebCore::MessagePort::isEntangled): Now returns false if the port has been closed. (WebCore::MessagePort::isCloned): Added new API to differentiate between cloned and closed ports (closed ports can still be passed to postMessage).
12:02 Changeset [51391] by zoltan@webkit.org

2009-11-25 Zoltan Horvath <zoltan@webkit.org>

Reviewed by Eric Seidel.

Change run_command to give back stderr by default
https://bugs.webkit.org/show_bug.cgi?id=31734

Change run_command to give back stderr by default.
Set run_commands's 'svn-create-patch' calling to put only the stdout into the patches.
Change the related unittest call.

  • Scripts/modules/scm.py:
  • Scripts/modules/scm_unittest.py:
10:22 Changeset [51390] by kov@webkit.org

Skip test that needs structure we do not provide yet

10:03 Changeset [51389] by dglazkov@chromium.org

2009-11-25 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Darin Fisher.

[Chromium] Plumb setNeedsSiteSpecificQuirks throught WebKit API.
https://bugs.webkit.org/show_bug.cgi?id=31878

  • public/WebSettings.h: Added decl.
  • src/WebSettingsImpl.cpp: (WebKit::WebSettingsImpl::setNeedsSiteSpecificQuirks): Added impl.
  • src/WebSettingsImpl.h:
09:12 Changeset [51388] by hausmann@webkit.org

[Qt] Fix crash of QtWebKit on any page with Flash when compiled with MinGW.

Patch by Jocelyn Turcotte <jocelyn.turcotte@nokia.com> on 2009-11-25
Reviewed by Simon Hausmann.

Fix inline assembly, don't dereference the function pointer twice.

  • plugins/win/PluginViewWin.cpp:

(WebCore::PluginView::hookedBeginPaint):
(WebCore::PluginView::hookedEndPaint):

09:02 Changeset [51387] by hausmann@webkit.org

[Qt] Remove the Referer header when redirecting to a non-secure site
https://bugs.webkit.org/show_bug.cgi?id=31785

Patch by Jakub Wieczorek <faw217@gmail.com> on 2009-11-22
Reviewed by Adam Barth.

This makes Qt pass two tests introduced in r50226.

WebCore:

  • platform/network/qt/QNetworkReplyHandler.cpp:

(WebCore::QNetworkReplyHandler::sendResponseIfNeeded):

LayoutTests:

  • platform/qt/Skipped:
08:50 Changeset [51386] by benm@google.com

The select elements do not reflect the actual choice the user makes.
https://bugs.webkit.org/show_bug.cgi?id=31831

Patch by Andrei Popescu <andreip@google.com> on 2009-11-25
Reviewed by Dimitri Glazkov.

Handle the drawing of the listboxes in Android code.

No new features, just fixing an Android problem. Existing layout tests are sufficient.

  • platform/android/RenderThemeAndroid.cpp:

(WebCore::theme):
(WebCore::RenderThemeAndroid::platformActiveSelectionBackgroundColor):
(WebCore::RenderThemeAndroid::platformActiveListBoxSelectionBackgroundColor):
(WebCore::RenderThemeAndroid::platformInactiveListBoxSelectionBackgroundColor):
(WebCore::RenderThemeAndroid::platformActiveListBoxSelectionForegroundColor):
(WebCore::RenderThemeAndroid::platformInactiveListBoxSelectionForegroundColor):
(WebCore::RenderThemeAndroid::adjustButtonStyle):
(WebCore::RenderThemeAndroid::paintTextArea):
(WebCore::RenderThemeAndroid::adjustListboxStyle):

  • platform/android/RenderThemeAndroid.h:
08:15 Changeset [51385] by eric@webkit.org

2009-11-25 Alexander Pavlov <apavlov@chromium.org>

Reviewed by Pavel Feldman.

Load InspectorResource mime type from CachedResource for 304 responses

For network libraries that do not merge cache data into 304 ResourceResponses,
the mime type for cached resources is unknown.
https://bugs.webkit.org/show_bug.cgi?id=31868

  • inspector/InspectorResource.cpp: (WebCore::InspectorResource::updateResponse): (WebCore::InspectorResource::updateScriptObject): (WebCore::InspectorResource::type):
08:06 Changeset [51384] by eric@webkit.org

2009-11-25 Andrei Popescu <andreip@google.com>

Reviewed by Eric Seidel.

Android is missing implementation of SSL Key generator functions.
https://bugs.webkit.org/show_bug.cgi?id=31825

This change adds the PlatformBridge class to platform/android.
PlarformBridge is used to access the embedding layer for things
such as key generator, cookies, plugins, etc.

No new tests required, this is platform code.

  • platform/android/PlatformBridge.h: Added.
  • platform/android/SSLKeyGeneratorAndroid.cpp: Added. (WebCore::getSupportedKeySizes): (WebCore::signedPublicKeyAndChallengeString):
08:05 Changeset [51383] by eric@webkit.org

2009-11-25 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Centralize required argument parsing in Command
https://bugs.webkit.org/show_bug.cgi?id=31872

  • Scripts/modules/commands/download.py: remove custom required arg message.
  • Scripts/modules/commands/upload.py: ditto.
  • Scripts/modules/multicommandtool.py:
    • Add _parse_required_arguments.
    • Pass program name off to OptionParser.
    • Add name() for access to tool name.
    • Add check_arguments_and_execute and make it return a return code.
    • Replace a couple uses of + with %.
  • Scripts/modules/multicommandtool_unittest.py: test _parse_required_arguments
07:59 Changeset [51382] by eric@webkit.org

2009-11-25 Philippe Normand <pnormand@igalia.com>

Reviewed by Eric Seidel.

[GTK] use gst_init_check() instead of gst_init()
https://bugs.webkit.org/show_bug.cgi?id=31864

Use gst_init_check() instead of gst_init() to prevent eventual
unexpected exit of the application.

  • platform/graphics/gtk/MediaPlayerPrivateGStreamer.cpp: (WebCore::do_gst_init): (WebCore::MediaPlayerPrivate::isAvailable):
07:46 Changeset [51381] by eric@webkit.org

2009-11-25 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Abstract out capturing stdout/stderr into a new OutputCapture class for re-use among the various unit tests.
https://bugs.webkit.org/show_bug.cgi?id=31870

  • Scripts/modules/commands/queries_unittest.py: Use the new class.
  • Scripts/modules/multicommandtool_unittest.py: Ditto.
  • Scripts/modules/outputcapture.py: Added.
05:01 Changeset [51380] by kov@webkit.org

Reviewed by Xan Lopez.

Some tests are crashing from time to time
https://bugs.webkit.org/show_bug.cgi?id=31866

Make sure we do not notify the client of a finished load, if the
load has been cancelled, or the client is gone.

  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::):
03:02 Changeset [51379] by benm@google.com

[Android] Upstream Android changes to WebCore/bridge/jni
https://bugs.webkit.org/show_bug.cgi?id=31824

Reviewed by Eric Seidel.

No new tests required as no new functionality.

  • bridge/jni/jni_class.cpp:

(JavaClass::JavaClass): Add calls to delete allocated references, to avoid potential leaks.

  • bridge/jni/jni_instance.cpp: Add an Android include path.
  • bridge/jni/jni_instance.h: Add getter/setter for JObjectWrapper::_instance and make the JavaInstance ctor and member variables protected. Both needed for the Android port, see bug for discussion.

(JSC::Bindings::JObjectWrapper::instance): Added.
(JSC::Bindings::JObjectWrapper::setInstance): Added.

  • bridge/jni/jni_runtime.cpp:

(JavaMethod::JavaMethod): Delete an allocated reference to avoid a potential leak.

  • bridge/jni/jni_utility.h:

(JSC::Bindings::callJNIMethodV): Delete an allocated reference to avoid a potential leak.

01:40 QtWebKitContrib edited by hausmann@webkit.org
(diff)
00:25 Changeset [51378] by zoltan@webkit.org

Allow custom memory allocation control for classes of the plugins and rendering directory in WebCore
https://bugs.webkit.org/show_bug.cgi?id=31827

Reviewed by Eric Seidel.

Inherits the following classes from Noncopyable because these are instantiated
by 'new' and no need to be copyable:

class/struct name - instantiated at: WebCore/'location'

class PluginRequest - plugins/PluginView.cpp:521
class PluginMainThreadScheduler - plugins/PluginMainThreadScheduler.cpp:34
class PluginDatabase - plugins/PluginDatabase.cpp:50
struct MimeClassInfo - plugins/PluginInfoStore.cpp:50
struct PluginInfo - plugins/PluginInfoStore.cpp:40
class RenderArena - dom/Document.cpp:1401
class RenderImageScaleData - rendering/RenderImage.cpp:149
class TableLayout - (its child class) rendering/RenderTable.cpp:82
struct ColumnInfo - rendering/RenderBlock.cpp:3590
struct FloatingObject - rendering/RenderBlock.cpp:2300
struct MaxMargin - rendering/RenderBlock.cpp:4794
class RenderMarquee - rendering/RenderLayer.cpp:3277

  • plugins/PluginData.h:
  • plugins/PluginDatabase.h:
  • plugins/PluginMainThreadScheduler.h:
  • plugins/PluginView.h:
  • rendering/RenderArena.h:
  • rendering/RenderBlock.cpp:
  • rendering/RenderBlock.h:
  • rendering/RenderImage.cpp:
  • rendering/RenderMarquee.h:
  • rendering/TableLayout.h:
Note: See TracTimeline for information about the timeline view.