Closed Bug 956105 Opened 10 years ago Closed 10 years ago

Crash in ~CxxStackFrame while running MO call scripts

Categories

(Core :: IPC, defect)

28 Branch
ARM
Gonk (Firefox OS)
defect
Not set
critical

Tracking

()

RESOLVED DUPLICATE of bug 957828
blocking-b2g 1.3+
Tracking Status
b2g-v1.3 --- affected

People

(Reporter: ggrisco, Unassigned)

Details

(Keywords: crash, regression, Whiteboard: [caf priority: p3][CR 593782][b2g-crash])

Crash Data

Attachments

(1 file)

Test Steps:
1. Run MO call test script.
2. After 2days of run, mini dumps are generated in the phone.

  [@ IPC::Message::is_interrupt() const | mozilla::ipc::MessageChannel::CxxStackFrame::~CxxStackFrame() | mozilla::ipc::MessageChannel::Send(IPC::Message*) | mozilla::dom::PContentParent::SendSetOffline(bool const&) ]
Component: General → IPC
Product: Firefox OS → Core
Version: unspecified → 28 Branch
Keywords: crash, regression
Whiteboard: [CR 593782] → [CR 593782][b2g-crash]
Crash Signature: [@ IPC::Message::is_interrupt() const]
Is there more to the stack trace? This is almost certainly use-after-free of PContentParent, and a bug in the caller not an IPC bug.
Flags: needinfo?(ggrisco)
Oh nevermind, I missed the attachment. Hrm. You can't really tell from the stack whether the PContentParent is dead (probably not, since it's refcounted). It's unlikely that the ~CxxStackFrame is dead. But CxxstackFrame.mThat could possibly have died in the middle of a call if the other side shut down in the middle of the call?
Flags: needinfo?(ggrisco) → needinfo?(dvander)
Andrew,

Can you please take a look and reassign?
Flags: needinfo?(overholt)
1.3+ as this blocks QC. Need a solution for the same by 1/10/14.
blocking-b2g: 1.3? → 1.3+
Blake and/or Ben can take a look.
Flags: needinfo?(overholt)
Flags: needinfo?(mrbkap)
Flags: needinfo?(bent.mozilla)
I haven't been able to identify anything obviously problematic with the code here, but I did stumble across bug 957828. It's possible that those patches could wallpaper over this bug, though I wouldn't want to be much money on it. Greg, can you try applying those patches and re-run your tests?
blocking-b2g: 1.3+ → 1.3?
Flags: needinfo?(bent.mozilla) → needinfo?(ggrisco)
Resetting blocking flag, as this still blocks the QC tracking bug.
blocking-b2g: 1.3? → 1.3+
(In reply to ben turner [:bent] (use the needinfo? flag!) from comment #6)
> I haven't been able to identify anything obviously problematic with the code
> here, but I did stumble across bug 957828. It's possible that those patches
> could wallpaper over this bug, though I wouldn't want to be much money on
> it. Greg, can you try applying those patches and re-run your tests?

Hi ben,

I will try fix from bug 957828 and update here soon.
moving ni to tapas
Flags: needinfo?(ggrisco) → needinfo?(tkundu)
I am still working with our internal team to get results from monkey testing.(In reply to ben turner [:bent] (use the needinfo? flag!) from comment #10)
> Any updates here?

I am still working with our internal team to get results from monkey testing with fix from bug 957828.

I will update here asap.
Sounds like we have a path going forward already, clearing ni request on me (and dvander who's unlikely to do anything here).
Flags: needinfo?(mrbkap)
Flags: needinfo?(dvander)
(In reply to Tapas Kumar Kundu from comment #11)
> I am still working with our internal team to get results from monkey
> testing.(In reply to ben turner [:bent] (use the needinfo? flag!) from
> comment #10)
> > Any updates here?
> 
> I am still working with our internal team to get results from monkey testing
> with fix from bug 957828.
> 
> I will update here asap.

Thanks. In order to make sure this bug has owner and it's on-going, do you think that you can be the assignee for this bug? Thank you.
(In reply to Kevin Hu [:khu] from comment #13)
> (In reply to Tapas Kumar Kundu from comment #11)
> > I am still working with our internal team to get results from monkey
> ...
> Thanks. In order to make sure this bug has owner and it's on-going, do you
> think that you can be the assignee for this bug? Thank you.

Sorry for late. This problem is resolved by fix from  bug 957828.
Flags: needinfo?(tkundu)
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → DUPLICATE
To be clear, my worry in comment 6 stands. I think this probably just wallpapered over some other problem :-/
Whiteboard: [CR 593782][b2g-crash] → [caf priority: p3][CR 593782][b2g-crash]
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: