Bug 136019

Summary: REGRESSION (r168244): Content in horizontal-bt page is offset such that only the end is viewable and there is a white gap at the top
Product: WebKit Reporter: mitz
Component: Layout and RenderingAssignee: Simon Fraser (smfr) <simon.fraser>
Severity: Normal CC: bdakin, cmarcelo, commit-queue, esprehn+autocc, glenn, jamesr, kondapallykalyan, luiz, simon.fraser, thorton, tonikitoo
Priority: P1 Keywords: InRadar, Regression
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Description Flags
Test case
Patch mitz: review+

Description mitz 2014-08-16 09:15:02 PDT
Created attachment 236717 [details]
Test case

To reproduce the bug, open the test case in Safari. The content is pushed down such that there is a white gap at the top and only the top (end) of the content is visible.
Comment 1 mitz 2014-08-16 09:15:34 PDT
Caused by http://trac.webkit.org/r168244, the fix for bug 132535.
Comment 2 mitz 2014-08-16 09:17:19 PDT
Comment 3 Simon Fraser (smfr) 2016-01-03 11:11:24 PST
Fixed, probably in https://trac.webkit.org/changeset/194487.
Comment 4 Simon Fraser (smfr) 2016-01-21 20:07:52 PST
Not fixed.
Comment 5 Simon Fraser (smfr) 2016-01-21 20:09:06 PST
The bug is that RLC sets the correct position of the root content layer at:
m_rootContentLayer->setPosition(FloatPoint(documentRect.x(), documentRect.y() + m_renderView.frameView().yPositionForRootContentLayer()));

but the scrolling thread overwrite that with a bad position at:
        m_scrolledContentsLayer.get().position = FloatPoint(m_scrolledContentsLayer.get().position.x,
            FrameView::yPositionForRootContentLayer(position, topContentInset, headerHeight()));

which ignores the documentRect part.
Comment 6 Simon Fraser (smfr) 2016-01-21 21:14:53 PST
Created attachment 269547 [details]
Comment 7 Simon Fraser (smfr) 2016-01-21 22:45:47 PST