Patch bug_1152046_prclose_v2.patch as push on 14.Sep. and we wanted to back it out because of a problems that i cause on linux, only a couple of hours later. Backing out did not work because crash described in https://bugzilla.mozilla.org/show_bug.cgi?id=1204837#c14 . Now because of a regression we wanted to disable it temporary on b2g, but it cause again the same problem. Bug 1152046 introduce a new thread that is only doing PR_Close calls. So all PR_Close call are not performed on socketThread as it use to be but on ClosingService thread. Patch bug_1152046_disable_on_ffos.patch disables ClosingService for b2g, so revert the behavior as it use to be before 14.Sep. I think https://bugzilla.mozilla.org/show_bug.cgi?id=1152046#c87 was nothing to do with ClosingService, but only because of timing change it exposes another bug.
Bug 1127270 is the last bug that touches ImageBridgeParent and it landed on the same day as the patch from 1152046. I assume that is the cause. What a coincidence.
It is possible that this has cause problem in bug 1207048. It has also happened after disabling ClosingService, but the change in ClosingService could not cause it.
I don't think bug 1127270 would cause this, Nical, do you have any ideas?
(In reply to Bas Schouten (:bas.schouten) from comment #4) > I don't think bug 1127270 would cause this, Nical, do you have any ideas? I don't see how bug 1127270 could have caused this either off hands but shutdown is full of mysteries and surprises, so it is hard to be 100% sure. There is very little information on this bug to act on, though. Is there a stack trace or something that provides us with any bit of information about the issue (other than that it is related to ImageBridgeParent)? Is it only on b2g?
Probably bug 1168337 had fixed this. We will see, since ClosingService will be removed (bug 1238017).
You need to log in before you can comment on or make changes to this bug.