| Summary: | ASSERTION FAILED: &formattingContextRoot == m_rootContainer.get() in WebCore::Layout::LayoutState::establishedInlineFormattingState | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Ryan Haddad <ryanhaddad> | ||||||
| Component: | Layout and Rendering | Assignee: | Simon Fraser (smfr) <simon.fraser> | ||||||
| Status: | RESOLVED FIXED | ||||||||
| Severity: | Normal | CC: | ap, bfulgham, cdumez, esprehn+autocc, ews-watchlist, kangil.han, simon.fraser, webkit-bot-watchers-bugzilla, webkit-bug-importer, zalan | ||||||
| Priority: | P2 | Keywords: | InRadar | ||||||
| Version: | WebKit Nightly Build | ||||||||
| Hardware: | Unspecified | ||||||||
| OS: | Unspecified | ||||||||
| Attachments: |
|
||||||||
|
Description
Ryan Haddad
2020-09-17 15:36:43 PDT
fast/layoutformattingcontext/table-basic-row-baseline-with-nested-table.html fast/layoutformattingcontext/table-flex-width-border-collapse.html https://build.webkit.org/results/Apple%20Mojave%20Debug%20WK2%20(Tests)/r267204%20(12039)/results.html Created attachment 409073 [details]
crash log
RuntimeEnabledFeatures::sharedFeatures().layoutFormattingContextIntegrationEnabled() is not supposed to be true in LayoutState::establishedInlineFormattingState() while running these test cases. fast/layoutformattingcontext/anonymous-inline-container-simple.html https://build.webkit.org/results/Apple-Catalina-Debug-WK2-Tests/r267166%20(6676)/results.html fast/layoutformattingcontext/absolute-positioned-simple-table3.html fast/layoutformattingcontext/table-fixed-width-with-max-distribution.html https://build.webkit.org/results/Apple-Catalina-Debug-WK2-Tests/r267141%20(6665)/results.html fast/layoutformattingcontext/absolute-positioned-simple-table2.html fast/layoutformattingcontext/table-cell-vertical-alignment-simple.html https://build.webkit.org/results/Apple-Catalina-Debug-WK2-Tests/r267206%20(6698)/results.html This appears to be a recent regression (maybe in the last 5 days?) and it is slowing down EWS. I think it's caused by https://trac.webkit.org/changeset/267024/webkit (Simon can confirm it). The patch is correct, it's just that something happens to the runtime switch while running/in-between the tests. We crash *after* loading about:blank between tests. I think about:blank has some special handling, so maybe we leave the LayoutState from the previous test around? Created attachment 409193 [details]
Patch
Committed r267310: <https://trac.webkit.org/changeset/267310> All reviewed patches have been landed. Closing bug and clearing flags on attachment 409193 [details]. |