Summary: | [Qt] 6 SVG tests assert after r107207 | ||
---|---|---|---|
Product: | WebKit | Reporter: | Csaba Osztrogonác <ossy> |
Component: | Tools / Tests | Assignee: | Nobody <webkit-unassigned> |
Status: | RESOLVED WORKSFORME | ||
Severity: | Critical | CC: | kkristof, ossy, rhodovan.u-szeged, zherczeg, zimmermann |
Priority: | P1 | Keywords: | Qt, QtTriaged |
Version: | 528+ (Nightly build) | ||
Hardware: | All | ||
OS: | All | ||
Bug Depends on: | 67543 | ||
Bug Blocks: | 78115, 78219, 79668 |
Description
Csaba Osztrogonác
2012-02-10 00:46:27 PST
I skipped the to paint the bot green - http://trac.webkit.org/changeset/107378 Please unskip them with the proper fix. > ASSERTION FAILED: m_transparencyCount > 0
Hmpf, the problem is the m_transparencyCount generalization. Qt maintains a different count "layerCount" returned by isInTransparentLayer(), and it calls endTransparencyLayer too often, compared to all other ports - to support image clipping.
Cairo has similar needs (mask image operation during restore()) but doesn't suffer from the problem - can this be reused instead? I hope Zoltan can have a look.
Bug 67543 is the culprit, leading to the Qt assertion. Two more asserting test: svg/dynamic-updates/SVGMaskElement-svgdom-width-prop.html svg/dynamic-updates/SVGMaskElement-dom-x-attr.html (In reply to comment #4) > Two more asserting test: > svg/dynamic-updates/SVGMaskElement-svgdom-width-prop.html > svg/dynamic-updates/SVGMaskElement-dom-x-attr.html And svg/dynamic-updates/SVGPathElement-dom-requiredFeatures.html I skipped them too. I hope once the test coverge will hit 0% and then all tests will pass forever. :) Committed r117040: <http://trac.webkit.org/changeset/117040> |