Closed Bug 894933 Opened 6 years ago Closed 6 years ago
crash in gfx3DMatrix::Can
Draw2D(gfx Matrix*) const
This bug was filed from the Socorro interface and is report bp-8c633977-4434-4e1d-b32b-be1992130717 . ============================================================= This is a somewhat easy crash to hit on a Samsung GT-S5360 ARMv6 phone. Every time I hit this I was interacting with the awesome screen, often typing. More reports at https://crash-stats.mozilla.com/report/list?product=FennecAndroid&signature=gfx3DMatrix%3A%3ACanDraw2D%28gfxMatrix*%29+const
Component: Graphics, Panning and Zooming → Graphics: Layers
Product: Firefox for Android → Core
Why should this track? Is it a regression?
It is an easy to reproduce crash on the phone. May be related to the other ARMv6 top crash bug 845867?
Milan, can we get an assignee?
tracking-fennec: ? → +
The 25 (nightly) crashes are all sigsegv, the 22 (release) ones are a mixture of sigill and sigsegv some of which are tagged as "heap corruption" if I'm reading it right.
BenWa, can you take a look?
Assignee: nobody → bgirard
If someone can start the ball with a regression window it would make this much easier to solve.
All of these are deferred composites. Perhaps we're not holding a reference and by the time the Runnable executes the object has been freed.
Attachment #8390741 - Flags: review?(jmuizelaar) → review+
https://hg.mozilla.org/integration/mozilla-inbound/rev/c06738b18379 Let's leave open until we can confirm. It's possible that we have more dangling stuff closer to the LayerManager.
Looks like the patch resolved the problem.
You need to log in before you can comment on or make changes to this bug.