[Webkit-unassigned] [Bug 254062] New: [Filters] Calculating the CSSFilter geometry and clipping sometimes is incorrect

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=254062

            Bug ID: 254062
           Summary: [Filters] Calculating the CSSFilter geometry and
                    clipping sometimes is incorrect
           Product: WebKit
           Version: WebKit Nightly Build
          Hardware: Unspecified
                OS: Unspecified
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: Layout and Rendering
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: sabouhallawa at apple.com
                CC: bfulgham at webkit.org, simon.fraser at apple.com,
                    zalan at apple.com

Like SVGFilter, CSSFilter should always intersect the FilterEffect imageRect with its primitiveSubregion. This will require:

1. Filter::ClipOperation will be removed.
2. Filter::clipToMaxEffectRect() will always clip the rectangle with maxEffectRect.
3. FilterEffect::calculateImageRect() will return the maxEffectRect if number of input effects is zero. This was a bug which was covered by the union mode of Filter::clipToMaxEffectRect().
4. RenderLayerFilters::beginFilterEffect() should set its targetBoundingBox the sourceImageRect of FilterTargetSwitcher.
5. SVGFilterGraph::getNamedNodes() should not return an error if one of the inputs is builtin and there is no node for it. This happens when we build a graph of primitives and one of the inputs is SourceGraphic or SourceAlpha. In this case we just need to ignore this input since there is no built-in primitive for it. This fixes the calculation of the filter outsets which was covered by  the union mode of Filter::clipToMaxEffectRect().

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20230317/64b83f65/attachment.htm>