GitLab

Biquad filters have a tail where zero input can produce non-zero
output due memory from previous inputs.  However, we were
disconnecting biquad filter nodes when the input is disconnected such
as when a source node is finished.  This causes the biquad to suddenly
produce zero output.

Add a case to handle biquad filters like convolver nodes and delay
nodes which don't disconnect their outputs even if the inputs are
disconnected.

This is a temporary solution until full tail-time supported is implemented.
That will fix this issue, and also allow biquad's to be collected when
no longer needed.  With this change, biquad's will be active for a
very long time even with no inputs.

BUG=521155
TEST=biquad-tail.html

Review URL: https://codereview.chromium.org/1285073005

git-svn-id: svn://svn.chromium.org/blink/trunk@200691 bbb929c8-8fbe-4397-9dbb-9b2b20218538

nextChild may be removed from the DOM tree during the
|parserRemoveChild(*newChild)| call which triggers unload events of newChild's 
descendant iframes. In order to maintain the integrity of the DOM tree, the
insertion of newChild must be aborted in this case.

This patch adds a return statement that rectifies the behavior in this
edge case.

BUG=519558

Review URL: https://codereview.chromium.org/1283263002

git-svn-id: svn://svn.chromium.org/blink/trunk@200690 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=501896
NOTRY=true

Review URL: https://codereview.chromium.org/1296013002

git-svn-id: svn://svn.chromium.org/blink/trunk@200689 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=439376
TBR=yurys@chromium.org, oilpan-reviews@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/1298843002

git-svn-id: svn://svn.chromium.org/blink/trunk@200688 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch makes files in "core/editing/" to match Blink coding style and
formatting comments, near style changes, to fit 80 characters to ease of working
these files in the future for improving code health.

Note: Naming related style violation will be addressed in other patches to
reduce size of patches.

This patch is also a preparation of moving execCommand related files,
http://crrev.com/1294543005

BUG=509257
TEST=n/a; no behavior changes

Review URL: https://codereview.chromium.org/1292303005

git-svn-id: svn://svn.chromium.org/blink/trunk@200687 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=519918

Review URL: https://codereview.chromium.org/1287423005

git-svn-id: svn://svn.chromium.org/blink/trunk@200686 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://src.chromium.org/viewvc/blink?view=revision&revision=200677

BUG=509025
TBR=leviw@chromium.org

Review URL: https://codereview.chromium.org/1287903005 .

git-svn-id: svn://svn.chromium.org/blink/trunk@200684 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This is output of webkit-patch print-flaky-tests.

TBR=leviw@chromium.org
BUG=521862, 521861, 521860, 521859, 521858, 521857, 521856, 521855, 521854, 521853, 521852, 521852, 521852, 521852

Review URL: https://codereview.chromium.org/1300733003 .

git-svn-id: svn://svn.chromium.org/blink/trunk@200682 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=515518
R=pfeldman@chromium.org, pfeldman

Review URL: https://codereview.chromium.org/1282593003 .

git-svn-id: svn://svn.chromium.org/blink/trunk@200681 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Re-rebaseline xp results after r200605 (patchset #1 id:1 of https://codereview.chromium.org/1296753004/ )

Reason for revert:
webkit_lint errors due to duplicate TestExpectation entries

http://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.6/builds/43492/steps/webkit_lint/logs/stdio

The test is already marked as failing by https://codereview.chromium.org/1286343003 Auto-rebase didn't help first time that's why I preferred skipping it on XP and let the author to take care of it.

Original issue's description:
> Re-rebaseline xp results after r200605
> 
> The rest of the bots updated and were happy. XP kept failing...
> 
> BUG=449197
> TBR=fmalita@chromium.org
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200676

TBR=fmalita@chromium.org,leviw@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=449197

Review URL: https://codereview.chromium.org/1299473004

git-svn-id: svn://svn.chromium.org/blink/trunk@200680 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=521704,496259

Review URL: https://codereview.chromium.org/1286373004

git-svn-id: svn://svn.chromium.org/blink/trunk@200679 bbb929c8-8fbe-4397-9dbb-9b2b20218538

InspectorDebuggerAgent now implements V8DebuggerAgent::Client and aggregates V8DebuggerAgent. The latter will be moved into a reusable component which cannot depend on Blink.

BUG=439376

Review URL: https://codereview.chromium.org/1286343003

git-svn-id: svn://svn.chromium.org/blink/trunk@200678 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The widths just changed by a pixel

BUG=509049
TBR=eae@chromium.org

Review URL: https://codereview.chromium.org/1252603003

git-svn-id: svn://svn.chromium.org/blink/trunk@200677 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The rest of the bots updated and were happy. XP kept failing...

BUG=449197
TBR=fmalita@chromium.org

Review URL: https://codereview.chromium.org/1296753004

git-svn-id: svn://svn.chromium.org/blink/trunk@200676 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=521801
TBR=leviw,fmalita@chromium.org,
NOTRY=true

Review URL: https://codereview.chromium.org/1298063002

git-svn-id: svn://svn.chromium.org/blink/trunk@200675 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=sergeyv@chromium.org, sergeyv

Review URL: https://codereview.chromium.org/1297053002 .

git-svn-id: svn://svn.chromium.org/blink/trunk@200674 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=520066
NOTRY=true

Review URL: https://codereview.chromium.org/1291673005

git-svn-id: svn://svn.chromium.org/blink/trunk@200673 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=sergeyv@chromium.org, pfeldman, sergeyv

Review URL: https://codereview.chromium.org/1296513005 .

git-svn-id: svn://svn.chromium.org/blink/trunk@200672 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Move dispatchDidFirstVisuallyNonEmptyLayout into WebViewImpl. (patchset #2 id:20001 of https://codereview.chromium.org/1295053002/ )

Reason for revert:
Looks like it caused a race in:

ProfileErrorBrowserTestInstance/ProfileErrorBrowserTest.CorruptedProfile


Original issue's description:
> Move dispatchDidFirstVisuallyNonEmptyLayout into WebViewImpl.
> 
> This removes the last user of didLayoutWithPendingStylesheets() and instead moves
> the call to dispatchDidFirstVisuallyNonEmptyLayout into WebViewImpl so the
> notification to the embedder is tied to the pumping of frames. This changes the
> behavior of this notification to only dispatch for the top level frame,
> but Chromium already has a check for that:
> 
> https://chromium.googlesource.com/chromium/src/+/976d4d9a2735bdc11e5a641c84b6382566d48f1d/content/renderer/render_frame_impl.cc#3472
> 
> RenderFrameImpl::didFirstVisuallyNonEmptyLayout's first check is
> 
> if (frame->parent())
>     return;
> 
> so we already ignore all frames that are not the main frame. Future patches will
> clean up this interface so it's obvious this really only happens for the main
> frame.
> 
> BUG=521692
> R=dglazkov@chromium.org
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200664

TBR=dglazkov@chromium.org,ojan@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=521692

Review URL: https://codereview.chromium.org/1294683004

git-svn-id: svn://svn.chromium.org/blink/trunk@200669 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Using this method makes the code intent more clear.

Review URL: https://codereview.chromium.org/1296543003

git-svn-id: svn://svn.chromium.org/blink/trunk@200667 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert "Revert "Reland 2: mac: Use a placeholder string for the family name of the system font." Branch 2485"

I had intended to revert the CL on branch 2485, but the CQ sent it straight to
trunk. Undoing the revert.

TBR=tkent@chromium.org, keishi@chromium.org, thakis@chromium.org
NOTRY=true
NOPRESUBMIT=true

Review URL: https://codereview.chromium.org/1292593003

git-svn-id: svn://svn.chromium.org/blink/trunk@200666 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert "Reland 2: mac: Use a placeholder string for the family name of the system font." Branch 2485

This reverts commit c05256796e62264bd6f477581b905a02105b0aa1.

BUG=521034
TBR=tkent@chromium.org, keishi@chromium.org, thakis@chromium.org
NOTRY=true
NOPRESUBMIT=true

Review URL: https://codereview.chromium.org/1299653004

git-svn-id: svn://svn.chromium.org/blink/trunk@200665 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This removes the last user of didLayoutWithPendingStylesheets() and instead moves
the call to dispatchDidFirstVisuallyNonEmptyLayout into WebViewImpl so the
notification to the embedder is tied to the pumping of frames. This changes the
behavior of this notification to only dispatch for the top level frame,
but Chromium already has a check for that:

https://chromium.googlesource.com/chromium/src/+/976d4d9a2735bdc11e5a641c84b6382566d48f1d/content/renderer/render_frame_impl.cc#3472

RenderFrameImpl::didFirstVisuallyNonEmptyLayout's first check is

if (frame->parent())
    return;

so we already ignore all frames that are not the main frame. Future patches will
clean up this interface so it's obvious this really only happens for the main
frame.

BUG=521692
R=dglazkov@chromium.org

Review URL: https://codereview.chromium.org/1295053002 .

git-svn-id: svn://svn.chromium.org/blink/trunk@200664 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Reland Call dispatcher's releaseRegistration when done with registrations (patchset #2 id:40001 of https://codereview.chromium.org/1290073003/ )

Reason for revert:
Speculatively reverting to fix crashes in http/tests/cache/cached-main-resource.html

http://test-results.appspot.com/dashboards/flakiness_dashboard.html#showExpectations=true&tests=http%2Ftests%2Fcache%2Fcached-main-resource.html

Original issue's description:
> Reland Call dispatcher's releaseRegistration when done with registrations
> 
> Reland of https://codereview.chromium.org/1279323002/
> 
> OilPan objects can be garbage collected at any time, so it's necessary to check the pointers.
> 
> BUG=502214
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200657

TBR=jsbell@chromium.org,jkarlin@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=502214

Review URL: https://codereview.chromium.org/1288893004

git-svn-id: svn://svn.chromium.org/blink/trunk@200663 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=521730

Review URL: https://codereview.chromium.org/1299743002

git-svn-id: svn://svn.chromium.org/blink/trunk@200662 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of 2D canvas: remain in deferred rendering mode with canvas to canvas drawImage (patchset #2 id:20001 of https://codereview.chromium.org/1288773005/ )

Reason for revert:
Appears to be triggering MSAN crashes: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=virtual%2Fdisplay_list_2d_canvas%2Ffast%2Fcanvas%2F2d.composite.globalAlpha.fillPath.html&testType=layout-tests

Original issue's description:
> 2D canvas: remain in deferred rendering mode with canvas to canvas drawImage
> 
> If the source canvas is not animated, no need to exit deferred rendering
> because there is no risk of memory bloat due to snapshotting.
> 
> BUG=521001
> TEST=telemetry tough_canvas_cases
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200654

TBR=senorblanco@chromium.org,junov@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=521001

Review URL: https://codereview.chromium.org/1295673003

git-svn-id: svn://svn.chromium.org/blink/trunk@200660 bbb929c8-8fbe-4397-9dbb-9b2b20218538

While fontWeight was protected using !fontWeightNeedsResolving in
order to not call fontWeightIsBold for lighter/bolder, baseFontWeight
was not, so add that logic.

BUG=518572

Review URL: https://codereview.chromium.org/1284313002

git-svn-id: svn://svn.chromium.org/blink/trunk@200659 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Reland #1: Add a short term solution to fix OSX 10.11 system font rendering problems. (patchset #2 id:20001 of https://codereview.chromium.org/1289823008/ )

Reason for revert:
Broke layout tests on Mac 10.10

http://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.10/builds/2551/steps/webkit_tests/logs/stdio

Original issue's description:
> Reland #1: Add a short term solution to fix OSX 10.11 system font rendering problems.
> 
> I forgot to update the test FontFamilyMatcherMacTest.YosemiteFontWeights.
> 
> > I landed a CL (https://codereview.chromium.org/1276003003/) which fixes system
> > font problems on OSX 10.9 and OSX 10.10. The CL also fixed MatchNSFontFamily()
> > to correctly return the system font on 10.11, but this causes glyph rendering
> > problems. In the short term, I'm adding logic to MatchNSFontFamily() to return
> > nil on 10.11, which was the original behavior before I landed that CL.
> >
> > BUG=521034
> > Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200618
> 
> BUG=521034
> TBR=tkent@chromium.org, keishi@chromium.org
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200653

TBR=tkent@chromium.org,keishi@chromium.org,erikchen@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=521034

Review URL: https://codereview.chromium.org/1299723002

git-svn-id: svn://svn.chromium.org/blink/trunk@200658 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Reland of https://codereview.chromium.org/1279323002/

OilPan objects can be garbage collected at any time, so it's necessary to check the pointers.

BUG=502214

Review URL: https://codereview.chromium.org/1290073003

git-svn-id: svn://svn.chromium.org/blink/trunk@200657 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Both IE and FF skip firing of mouse transition events at deleted nodes,
with only IE doing it in a perfect manner by maintining the pairing of
entry/exit events. A crack in our code had been causing firing of these
events at a node after the node got deleted. This CL fixes the bug by
making chromium behave similar to FF. The perfect solution seems
non-trivial.

BUG=515921

Review URL: https://codereview.chromium.org/1288483003

git-svn-id: svn://svn.chromium.org/blink/trunk@200656 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=pfeldman
NOTRY=true

Review URL: https://codereview.chromium.org/1284003006

git-svn-id: svn://svn.chromium.org/blink/trunk@200655 bbb929c8-8fbe-4397-9dbb-9b2b20218538

If the source canvas is not animated, no need to exit deferred rendering
because there is no risk of memory bloat due to snapshotting.

BUG=521001
TEST=telemetry tough_canvas_cases

Review URL: https://codereview.chromium.org/1288773005

git-svn-id: svn://svn.chromium.org/blink/trunk@200654 bbb929c8-8fbe-4397-9dbb-9b2b20218538

I forgot to update the test FontFamilyMatcherMacTest.YosemiteFontWeights.

> I landed a CL (https://codereview.chromium.org/1276003003/) which fixes system
> font problems on OSX 10.9 and OSX 10.10. The CL also fixed MatchNSFontFamily()
> to correctly return the system font on 10.11, but this causes glyph rendering
> problems. In the short term, I'm adding logic to MatchNSFontFamily() to return
> nil on 10.11, which was the original behavior before I landed that CL.
>
> BUG=521034
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200618

BUG=521034
TBR=tkent@chromium.org, keishi@chromium.org

Review URL: https://codereview.chromium.org/1289823008

git-svn-id: svn://svn.chromium.org/blink/trunk@200653 bbb929c8-8fbe-4397-9dbb-9b2b20218538

If scroll is being restored we will still restore
both scroll and scale together otherwise we restore
scale on its own.

Here is the rationale of decoupling scale and scroll restoration:
https://groups.google.com/a/chromium.org/forum/#!msg/blink-dev/p1sy9aQDmtU/jndSickDCAAJ


BUG=444094

Review URL: https://codereview.chromium.org/1277583003

git-svn-id: svn://svn.chromium.org/blink/trunk@200652 bbb929c8-8fbe-4397-9dbb-9b2b20218538

In order to allow WebView and other projects to turn off the API if they
can't fully support it.

BUG=521319

Review URL: https://codereview.chromium.org/1299663002

git-svn-id: svn://svn.chromium.org/blink/trunk@200651 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Allowing PresentationRequest.start() to be called without user gesture
might lead to a confusion regarding the UI: it should show a picker and
usually such flow have to be started with a user gesture.

BUG=521590

Review URL: https://codereview.chromium.org/1296943003

git-svn-id: svn://svn.chromium.org/blink/trunk@200650 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This CL is to add a new interface and we are discussing as follows.

Intent to Implement and Ship : 
https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/SbQzldr3aOM

R=rbyers@chromium.org
BUG=498504

Review URL: https://codereview.chromium.org/1232003009

git-svn-id: svn://svn.chromium.org/blink/trunk@200649 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The function in question now returns a LineLayoutBoxModel instead of a
LayoutBoxModelObject*. To keep the CL size down, I've also added a
deprecatedBoxModelObject() method which still returns a LayoutBoxModelObject*,
because there are several references in core/paint/InlineFlowBoxPainter.cpp
that take the returned object and pass it around and call a bunch of methods
that aren't part of the line layout API yet. I will tackle those in a followup
patch.

BUG=499321

Review URL: https://codereview.chromium.org/1294483002

git-svn-id: svn://svn.chromium.org/blink/trunk@200647 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Add a short term solution to fix OSX 10.11 system font rendering problems. (patchset #2 id:20001 of https://codereview.chromium.org/1284993004/ )

Reason for revert:
Newly added test has been failing since its introduction

http://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=blink_platform_unittests&showExpectations=true&tests=FontFamilyMatcherMacTest.YosemiteFontWeights

Original issue's description:
> Add a short term solution to fix OSX 10.11 system font rendering problems.
> 
> I landed a CL (https://codereview.chromium.org/1276003003/) which fixes system
> font problems on OSX 10.9 and OSX 10.10. The CL also fixed MatchNSFontFamily()
> to correctly return the system font on 10.11, but this causes glyph rendering
> problems. In the short term, I'm adding logic to MatchNSFontFamily() to return
> nil on 10.11, which was the original behavior before I landed that CL.
> 
> BUG=521034
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=200618

TBR=tkent@chromium.org,keishi@chromium.org,erikchen@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=521034

Review URL: https://codereview.chromium.org/1296243002

git-svn-id: svn://svn.chromium.org/blink/trunk@200646 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Reentrance to WebViewImpl::closePagePopup() was possible.  closePopup() crashed
in that case because m_page was valid and m_page->mainFrame() was nullptr.

The details:
If WebViewImpl::closePagePopup() is called, and the LocalFrame in the page popup
has the last references to the page popup owner Element,
 1. WebPagePopupImpl::closePopup() calls destroyPage().
 2. destroyPage() calls Page::willBeDestroyed().
 3. willBeDestroyed() destructs the LocalFrame.
 4. The LocalFrame destructor destructs the owner Element.
 5. The owner Element destructor destructs PickerIndicatorElement.
 6. PickerIndicatorElement destructor calls WebViewImpl::closePagePopup().

This CL changes the code so that the frame doesn't have the last reference
to the owner Element.  We explicitly close the popup on
 - Owner detach (not PickerIndicatorElement detach)
 - Owner removal from the document tree.

This fixes a testcase in crbug.com/454043.  However we failed to make
a stable test.

BUG=454043

Review URL: https://codereview.chromium.org/1293953003

git-svn-id: svn://svn.chromium.org/blink/trunk@200645 bbb929c8-8fbe-4397-9dbb-9b2b20218538