Created attachment 8438576 [details] Crash Stack I'm seeing this reliably in a relatively unused profile (the one I use for testing WebRTC), although I've been unable to isolate it to any particular aspect of that profile. It happens with 100% reproducibility immediately upon start-up with this profile. Stack attached. If you need me to bring this up in a debugger so you can poke around at state, let me know.
Created attachment 8444027 [details] gdb output I just hit this as well, while loading three articles in tabs from http://www.nytimes.com/ (at least 2 of which were about today's World Cup matches).
nsSVGMaskFrame is in the attached stack, so basically we want to fix bug 987194 to fix this.
Depends on: 987194
It's in abr's stack and mcmanus's but not mine.
Can you attach yours?
Oh, you already did. Didn't bug 932198 fix that crash stack?
It certainly could have, given: $ hg log -r'ancestor(a6a457fe2a2a, 617e3d552045)' changeset: 188731:a6a457fe2a2a (i.e., that my crash was on a revision without bug 932198). Is bug 987194 going to happen soon? It seems bad to leave trunk in a state where certain combinations of Web features yield fatal assertions (and presumably worse things in non-DEBUG builds?); would it be possible to handle the relevant case instead?
It depends how quickly Matt gets to the review (he's been getting through my review request pretty quickly for other Moz2D conversion stuff). We should then get approval for aurora too.
Fixed by the patch for bug 987194.
Status: NEW → RESOLVED
Last Resolved: 4 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.