Closed
Bug 577740
Opened 14 years ago
Closed 14 years ago
Crash [@ mozilla::ipc::RPCChannel::CxxStackFrame::CxxStackFrame]
Categories
(Core :: IPC, defect)
Tracking
()
RESOLVED
WORKSFORME
People
(Reporter: mfinkle, Assigned: jdm)
Details
(Keywords: crash)
Crash Data
Attachments
(1 file, 1 obsolete file)
2.12 KB,
patch
|
Details | Diff | Splinter Review |
We see a lot of crashes in Fennec running on Maemo. See: http://crash-stats.mozilla.com/report/list?range_value=2&range_unit=weeks&signature=mozilla%3A%3Aipc%3A%3ARPCChannel%3A%3ACxxStackFrame%3A%3ACxxStackFrame&version=Fennec%3A2.0a1pre
Assignee | ||
Comment 1•14 years ago
|
||
Oooh, this is a fun one. Looks like something is invoking some JS on another thread which ends up calling nsIOService::SetOffline which invokes the various observers, so the ContentProcessParent tries to do IPDL stuff on the wrong thread. We could either just do a thread check in ContentProcessParent and potentially dispatch a runnable, or we could make the observer object an proxy. Decisions, decisions...
Assignee | ||
Comment 2•14 years ago
|
||
I don't believe I have any way of reproducing this crash, seeing as I don't have the hardware required. If somebody could confirm/deny whether this patch does anything, I'd be mighty appreciative.
Assignee: nobody → josh
Assignee | ||
Comment 3•14 years ago
|
||
Now with actual patch attached.
Attachment #456817 -
Attachment is obsolete: true
Assignee | ||
Comment 4•14 years ago
|
||
Actually, looking at the crash stacks some more, I might be totally off base about my earlier guess. If there's some way that somebody getting these crashes could reproduce them under GDB, I would be very interested in the results of
> fr 3
> p *this
Assignee | ||
Comment 5•14 years ago
|
||
Seeing as there have been no crashes with this stack since July 20th, I'm hopeful that this has gone away as a side-effect of some other landing.
OS: Mac OS X → Windows 7
Comment 6•14 years ago
|
||
if NS_IPC_IOSERVICE_SET_OFFLINE_TOPIC is being dispatch on the wrong thread, that would be bad. :( Either way, I don't think it is a bad thing to dispatch another event that actually does the call to SendSetOffline.
Assignee | ||
Comment 7•14 years ago
|
||
Haven't seen any of these crashes for a very long while; I'm going to mark this WFM.
Status: NEW → RESOLVED
Closed: 14 years ago
Resolution: --- → WORKSFORME
Updated•13 years ago
|
Crash Signature: [@ mozilla::ipc::RPCChannel::CxxStackFrame::CxxStackFrame]
You need to log in
before you can comment on or make changes to this bug.
Description
•