1034527 - SuspendThread is blocked on emulator

Status: NEW

(Toolkit :: Crash Reporting, defect)

Description

[StevenLee[:slee]]

As bug 1016184 comment 34, b2g may be blocked in SuspendThread when content process crashes.

Comment 1

[howie [:howie]]

Hi Mike, could you take a look for this issue? It blocks the copy/paste - a focus feature of Firefox OS. Thank you!

Comment 2

[Mike Hommey [:glandium]]

As far as I can tell from bug 1016184 comment 34, the problem is figured. Someone has to fix it. I currently don't have the time to be that someone.

Comment 3

[howie [:howie]]

Thanks Mike. Hi Taras, I know the resource must be tight here, but could you try to find someone to fix this issue? That would be really helpful for Firefox OS if you could, many thanks! :)

Comment 4

[(dormant account)]

Sorry, Mike is my only guy who can do this kind of stuff. Nick might have someone.

Comment 5

[Mike Hommey [:glandium]]

How about yoric?

Comment 6

[Nicholas Nethercote [inactive]]

I pointed erahm at this bug, but he has a number of other things on his plate at the moment and I don't know if he'll have time. How important is this? Is it just blocking the enabling of tests, or does it cause problems in normal use as well?

Comment 7

[StevenLee[:slee]]

(In reply to Nicholas Nethercote [:njn] from comment #6)
> I pointed erahm at this bug, but he has a number of other things on his
> plate at the moment and I don't know if he'll have time. How important is
> this? Is it just blocking the enabling of tests, or does it cause problems
> in normal use as well?
It seems that when the content process hits assertion when shutting down. Parent process may encounter this problem. I am not sure how many people have this problem. But it always happens when I turn on "touchcaret.enabled" and run the tests in gecko/layout/forms/test on emulator. I think it relates with some timing. If I attach the process with gdb, it does not always happen. After bug 774388 landed, it does not happen(because the content process does not hit the assertion). If someone has time, I think it's a good chance to solve it.

Comment 8

[Eric Rahm [:erahm]]

(In reply to StevenLee[:slee] from comment #7)
> It seems that when the content process hits assertion when shutting down.
> Parent process may encounter this problem. I am not sure how many people
> have this problem. But it always happens when I turn on "touchcaret.enabled"
> and run the tests in gecko/layout/forms/test on emulator. I think it relates
> with some timing. If I attach the process with gdb, it does not always
> happen. After bug 774388 landed, it does not happen(because the content
> process does not hit the assertion). If someone has time, I think it's a
> good chance to solve it.

So to sum up: this is no longer reproducible, but you are worried that it might show up again, correct?

Comment 9

[StevenLee[:slee]]

(In reply to Eric Rahm [:erahm] from comment #8)
> So to sum up: this is no longer reproducible, but you are worried that it
> might show up again, correct?
Yes. :)