Title: [140625] branches/chromium/1391a
- Revision
- 140625
- Author
- [email protected]
- Date
- 2013-01-23 18:14:15 -0800 (Wed, 23 Jan 2013)
Log Message
Revert 140206
Modified Paths
Removed Paths
Diff
Modified: branches/chromium/1391a/LayoutTests/ChangeLog (140624 => 140625)
--- branches/chromium/1391a/LayoutTests/ChangeLog 2013-01-24 02:11:18 UTC (rev 140624)
+++ branches/chromium/1391a/LayoutTests/ChangeLog 2013-01-24 02:14:15 UTC (rev 140625)
@@ -1,50 +1,3 @@
-2013-01-18 Alpha Lam <[email protected]>
-
- [chromium] Update test expectations
-
- Unreviewed. Update test expectations to mark test as flaky.
-
- * platform/chromium/TestExpectations:
-
-2013-01-18 Alpha Lam <[email protected]>
-
- [chromium] Update test expectations
- https://bugs.webkit.org/show_bug.cgi?id=107328
-
- Not reviewed. Mark this test as MISSING:
- http/tests/cache/cancel-multiple-post-xhrs.html
-
- * platform/chromium/TestExpectations:
-
-2013-01-18 Robert Hogan <[email protected]>
-
- fast/inline/fixed-pos-moves-with-abspos-parent-relative-ancestor.html is crashing after r140024
- https://bugs.webkit.org/show_bug.cgi?id=107209
-
- Reviewed by David Hyatt.
-
- * platform/chromium/TestExpectations:
-
-2013-01-18 Alpha Lam <[email protected]>
-
- [chromium] Layout test rebaselines
-
- Unreviewed rebaseline. Update test results after Chromium roll in r140190.
-
- * platform/chromium-win/css2.1/t100801-c544-valgn-03-d-agi-expected.png:
- * platform/chromium-win/css2.1/t100801-c544-valgn-03-d-agi-expected.txt:
- * platform/chromium/css2.1/t100801-c544-valgn-03-d-agi-expected.txt: Renamed from LayoutTests/platform/chromium-mac/css2.1/t100801-c544-valgn-03-d-agi-expected.txt.
-
-2013-01-18 Abhishek Arya <[email protected]>
-
- Heap-use-after-free in WebCore::RenderObject::isDescendantOf
- https://bugs.webkit.org/show_bug.cgi?id=107226
-
- Reviewed by David Hyatt.
-
- * fast/block/float/overhanging-float-not-removed-crash-expected.txt: Added.
- * fast/block/float/overhanging-float-not-removed-crash.html: Added.
-
2013-01-18 Chris Hopman <[email protected]>
Fix scrollRectToVisible in the presence of transforms
Deleted: branches/chromium/1391a/LayoutTests/fast/block/float/overhanging-float-not-removed-crash-expected.txt (140624 => 140625)
--- branches/chromium/1391a/LayoutTests/fast/block/float/overhanging-float-not-removed-crash-expected.txt 2013-01-24 02:11:18 UTC (rev 140624)
+++ branches/chromium/1391a/LayoutTests/fast/block/float/overhanging-float-not-removed-crash-expected.txt 2013-01-24 02:14:15 UTC (rev 140625)
@@ -1,2 +0,0 @@
-Bug 107226. Heap-use-after-free in WebCore::RenderObject::isDescendantOf.
-PASS. WebKit didn't crash
Deleted: branches/chromium/1391a/LayoutTests/fast/block/float/overhanging-float-not-removed-crash.html (140624 => 140625)
--- branches/chromium/1391a/LayoutTests/fast/block/float/overhanging-float-not-removed-crash.html 2013-01-24 02:11:18 UTC (rev 140624)
+++ branches/chromium/1391a/LayoutTests/fast/block/float/overhanging-float-not-removed-crash.html 2013-01-24 02:14:15 UTC (rev 140625)
@@ -1,33 +0,0 @@
-<!DOCTYPE html>
-<html>
-<noembed id=test4></noembed>
-<h6 id=test3>
-<h3 id=test7></h3>
-<span id=tCF36>
-<math>
-<and>
-<equivalent>
-<csc id=test5 style="float: left; "> X6DDguzp|m_v c;</csc>
-<otherwise id=test1>
-</equivalent>
-</and>
-<cartesianproduct id=test2>#L }|2pqynn</cartesianproduct>
-<factorial></factorial>
-<product id=test6>.
-</math>
-</span>
-</h6>
-<script>
-if (window.testRunner)
- testRunner.dumpAsText();
-
-document.documentElement.offsetTop;
-test1.appendChild(test2);
-test2.appendChild(test3);
-document.documentElement.offsetTop;
-test4.appendChild(test5);
-test6.appendChild(test7);
-document.documentElement.offsetTop;
-document.documentElement.innerHTML = "Bug 107226. Heap-use-after-free in WebCore::RenderObject::isDescendantOf.<br />PASS. WebKit didn't crash";
-</script>
-</html>
Modified: branches/chromium/1391a/Source/WebCore/ChangeLog (140624 => 140625)
--- branches/chromium/1391a/Source/WebCore/ChangeLog 2013-01-24 02:11:18 UTC (rev 140624)
+++ branches/chromium/1391a/Source/WebCore/ChangeLog 2013-01-24 02:14:15 UTC (rev 140625)
@@ -1,149 +1,3 @@
-2013-01-18 Brady Eidson <[email protected]>
-
- Rework NetworkProcess resource load identifiers.
- <rdar://problem/12934449> and https://bugs.webkit.org/show_bug.cgi?id=107192
-
- Reviewed by Alexey Proskuryakov.
-
- No new tests (No effect in tested configs).
-
- * WebCore.exp.in:
-
- Include an identifier argument for synchronous loads:
- * loader/LoaderStrategy.cpp:
- (WebCore::LoaderStrategy::loadResourceSynchronously):
- * loader/LoaderStrategy.h:
-
- Pass along a unique identifier to the synchronous load:
- * loader/FrameLoader.cpp:
- (WebCore::FrameLoader::loadResourceSynchronously):
-
- Remove the "setIdentifier" method:
- * loader/ResourceLoader.cpp:
- * loader/ResourceLoader.h:
-
-2013-01-18 Shinya Kawanaka <[email protected]>
-
- Introduce isHTMLTextFormControlElement and toHTMLTextFormControlElement instead of toTextFormControl
- https://bugs.webkit.org/show_bug.cgi?id=107089
-
- Reviewed by Kent Tamura.
-
- toTextFormControl(node) returns 0 if node is not HTMLTextFormControl. In recent coding convention,
- we only have ASSERT and statc_cast in this kind of conversion function. So this code looks like
- having a bad cast, though it does not.
-
- In this patch, we convert toTextFromControl to isHTMLTextFormControlElement and toHTMLFormControlElement,
- which aligns our coding convention.
-
- No new tests, simple refactoring.
-
- * editing/Editor.cpp:
- (WebCore::Editor::selectionForCommand):
- (WebCore::Editor::setBaseWritingDirection):
- * html/HTMLElement.cpp:
- (WebCore::HTMLElement::directionality):
- * html/HTMLTextFormControlElement.cpp:
- (WebCore::enclosingTextFormControl):
- * html/HTMLTextFormControlElement.h:
- (WebCore::isHTMLTextFormControlElement):
- (WebCore):
- (WebCore::toHTMLTextFormControlElement):
- * rendering/RenderTextControl.cpp:
- (WebCore::RenderTextControl::RenderTextControl):
- (WebCore::RenderTextControl::layoutSpecialExcludedChild):
- * testing/Internals.cpp:
- (WebCore::Internals::visiblePlaceholder):
-
-2013-01-18 Emil A Eklund <[email protected]>
-
- Change RenderStyle::getRoundedInnerBorderFor to take ints for border widths
- https://bugs.webkit.org/show_bug.cgi?id=107331
-
- Reviewed by Levi Weintraub.
-
- Borders are always represented as ints yet RenderStyle::
- getRoundedInnerBorderFor upcasts from int to LayoutUnit when
- calling getRoundedInnerBorderFor.
-
- No new tests, no change in functionality.
-
- * rendering/style/RenderStyle.cpp:
- (WebCore::RenderStyle::getRoundedInnerBorderFor):
- * rendering/style/RenderStyle.h:
- Change getRoundedInnerBorderFor to use ints for border widths.
-
-2013-01-18 Adam Barth <[email protected]>
-
- BackgroundHTMLParser shouldn't create AtomicStrings
- https://bugs.webkit.org/show_bug.cgi?id=107150
-
- Reviewed by Tony Gentilcore.
-
- Previously, we were creating an AtomicString from the HTMLToken for
- every start and end tag. This caused us to have a static dependency on
- the parser thread's atomic string table. We want to avoid static
- dependencies because we want to eventually run the background parser
- using a library like libdispatch that will let us run many version in
- parallel.
-
- This patch uses a different approach whereby we use the tag name from
- the CompactHTMLToken. In order to avoid calling strcmp excessively, we
- use StringImpl::hash to quickly reject matches with some common tag
- names. As a side effect, the hash value will be cached in the
- StringImpl, speeding up atomization once these StringImpls arrive on
- the main thread.
-
- This patch also improves the performance on html-parser-srcdoc.html.
- With the patch in bug 107236 applied (which focuses the benchmark more
- tightly on the parser), we spend 2.3% less time on the main thread
- (likely due to pre-caching the hash values) and 11.5% less time on the
- parser thread.
-
- * html/parser/BackgroundHTMLParser.cpp:
- (WebCore):
- (WebCore::threadSafeEqual):
- (WebCore::threadSafeMatch):
- (WebCore::BackgroundHTMLParser::pumpTokenizer):
-
-2013-01-18 Tony Gentilcore <[email protected]>
-
- Run defer scripts in the threaded HTML parser
- https://bugs.webkit.org/show_bug.cgi?id=107320
-
- Reviewed by Eric Seidel.
-
- No new tests because covered by http/tests/misc/script-defer.html.
-
- * html/parser/HTMLDocumentParser.cpp:
- (WebCore::HTMLDocumentParser::prepareToStopParsing):
- (WebCore::HTMLDocumentParser::didReceiveTokensFromBackgroundParser):
-
-2013-01-18 Robert Hogan <[email protected]>
-
- fast/inline/fixed-pos-moves-with-abspos-parent-relative-ancestor.html is crashing after r140024
- https://bugs.webkit.org/show_bug.cgi?id=107209
-
- Reviewed by David Hyatt.
-
- * rendering/RenderBlock.cpp:
- (WebCore::RenderBlock::layoutPositionedObjects): r140024 forgot to layout the fixedpos object after
- marking it.
-
-2013-01-18 Abhishek Arya <[email protected]>
-
- Heap-use-after-free in WebCore::RenderObject::isDescendantOf
- https://bugs.webkit.org/show_bug.cgi?id=107226
-
- Reviewed by David Hyatt.
-
- Test: fast/block/float/overhanging-float-not-removed-crash.html
-
- * rendering/RenderBox.cpp:
- (WebCore::RenderBox::removeFloatingOrPositionedChildFromBlockLists):
- Skip anonymous blocks in the chain to get the enclosing block and
- be able to correctly mark the overhanging floats in the next siblings.
-
2013-01-18 Chris Hopman <[email protected]>
Fix scrollRectToVisible in the presence of transforms
Modified: branches/chromium/1391a/Source/WebCore/rendering/RenderBox.cpp (140624 => 140625)
--- branches/chromium/1391a/Source/WebCore/rendering/RenderBox.cpp 2013-01-24 02:11:18 UTC (rev 140624)
+++ branches/chromium/1391a/Source/WebCore/rendering/RenderBox.cpp 2013-01-24 02:14:15 UTC (rev 140625)
@@ -179,12 +179,6 @@
}
if (parentBlock) {
- // Need to skip anonymous blocks in our ancestor chain since our overhanging floats
- // can be in the next siblings of enclosing block.
- while (parentBlock && parentBlock->isAnonymousBlock())
- parentBlock = parentBlock->containingBlock();
- ASSERT(parentBlock);
-
RenderObject* parent = parentBlock->parent();
if (parent && parent->isFlexibleBoxIncludingDeprecated())
parentBlock = toRenderBlock(parent);
_______________________________________________
webkit-changes mailing list
[email protected]
http://lists.webkit.org/mailman/listinfo/webkit-changes
