The patch for bug 997101 has resulted in layout/reftests/svg/svg-integration/filter-html-01-extref.xhtml failing intermittently. That patch simply made us stop setting an arbitrary default CSS viewport which causes a reflow before the reflow that happens when the size of the CSS viewport is subsequently set to the actual size of the device. So it seems likely that we have just exposed a race condition rather than there being anything wrong with bug 997101's patch. (It also exposed another race that was fixed in bug 1063073.) When considering the tradeoff of the performance benefit of bug 997101 for Firefox OS vs the exposure of the filter-html-01-extref.xhtml race condition for some SVGs, I think the type of SVG that is being hurt by this race (the type of thing that filter-html-01-extref.xhtml tests) is relatively rare. I think it's better to keep the bug 997101 fix to improve Firefox OS performance and work on figuring out and fixing the SVG race when we can.
I marked the test orange for B2G in: https://hg.mozilla.org/integration/mozilla-inbound/rev/d8155d385961
status-b2g-v2.1: --- → disabled
status-b2g-v2.2: --- → disabled
--> WONTFIX per bug 1306391. (side note: the test-annotation & comment that landed in comment 1 are being reverted in bug 1307332.)
Status: NEW → RESOLVED
Last Resolved: 2 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.