[Webkit-unassigned] [Bug 74196] [chromium] Delegate scroll events to the main thread when needed

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=74196


Sami Kyostila <skyostil at google.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #119033|review?                     |review-
               Flag|                            |




--- Comment #4 from Sami Kyostila <skyostil at google.com>  2011-12-14 12:53:02 PST ---
(From update of attachment 119033)
View in context: https://bugs.webkit.org/attachment.cgi?id=119033&action=review

Setting as review- until the preceding patch layer scrolling is updated.

>> Source/WebCore/platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:341

> 
> I know you're just moving code here, but I don't follow this logic and why the scroll layer behaves differently.  Would you mind explaining why this is? Don't all layers (root or no) have the page scale applied to them?
> 
> Also, do you need to apply the page scale delta? Maybe I'd feel more comfortable if there were more tests for this function that covered page scale.

Turns out the code here is not entirely correct, so I'll rework it in the preceding patch. Actually there is no need to differentiate between layers or scale the content point, but instead visibleLayerRect() must be scaled with the inverse of the page scale to get unscaled css coordinates that match the coordinate system of contentPoint.

The reason why page scale delta is not needed is that while we are zooming, visibleLayerRect() is recomputed in terms of the original page scale. It never includes m_pageScaleDelta so there is no need to undo that here. I'm not really sure whether this is intended, though.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.