05-23-2013 08:31 AM
Sorry I cannot agree to that. My Z10 uses 10.1 and still do unexpected reboots. Many of my colleagues have the same issue. It doesn't matter if it is 10.1 or 10.0 firmware.
What I miss here is an official statement from RIM. Could a RIM technical guy give a statement to that error?
05-23-2013 09:00 AM
BlackBerry is aware of this; there's a formal ticket on it in their system. 10.1.0.273 (the common release floating around for EU and Canadian carriers) does not fully address the problem, but does materially reduce it in my experience.
I've done a lot of investigation on this issue and have a decent operating theory on what's going on. My best guess is that it will be gone with the next formal release and likely will be available in a "leak" form before it formally hits the carriers.
05-23-2013 09:18 AM
I cannot speak to it with certainty but I can speak to what's very likely going on (being both a kernel programmer on Unix-style systems going back to the 1980s and having run a lot of tests on this problem myself -- while I can't quite get to "reproduce on demand" I'm pretty close to it.)
Here's my take on it.....
The problem is related to a deadlock and the reboot itself is almost-certainly coming from the watchdog timer, which is there for the explicit purpose of prohibiting a deadlock from freezing the device (which would require a battery pull.)
I suspected originally the radio code was likely implicated due to the wide variation of incidence, from many people having NO problems to others having fairly-easily-reproduced ones. Then we got the ability to swap radio versions (along with firmware) non-destructively and now I'm essentially convinced this is the case.
In addition there is symptomology that is broadly congruent with this operating theory related to the VPN and WiFi functionality not always coming back up. You will probably never see this happen to you unless you nail the VPN connection with autoconnect on both WiFi and Cell services, especially on leaked firmware.
It's a problem but it's a software one, is tickled by specific sets of events, and if I'm correct in what's going on it's going to be loads of fun to eliminate it simply because getting the exact state of the system when the deadlock occurs is critical to understanding exactly what sequence causes it.
Deadlocks happen when you have two exclusive-access operations that "cross" each other. The first takes a lock on a resource and starts working, then gets preempted for some reason. A second operation in a different thread takes a lock on a second resource and then attempts to obtain one on the first and blocks doing so as the first thread or process has locked that resource already. If the first process or thread is then scheduled for execution and tries to lock what the second process has acquired you're screwed as it will block as well. Now you have to processes or threads blocked and neither can proceed.
This is consistent with what I can reproduce with the VPN stack; it will refuse to come up under certain reselect conditions but the other end has authenticated and inserted the SPI and routing entries -- that is, the server says the link is up and good. Once it happens you have to tear down the underlying resource (in this case, by putting the unit in airplane mode) to clear it, after which the VPN will restart, because the first process that caused the deadlock releases it locks and by doing so the condition clears.
IF the process that takes the deadlock originally is running with elevated priority such that it always gets reselected next to run then the deadlock never clears. In this case the process that resets the watchdog timer never executes and when the timer expires the phone is reset.
I'll lay odds this is a purely software issue and it is centered in the radio code. I also have reason to believe BlackBerry knows where it is in general terms, because the behavior has changed materially in the various leaks, especially when the autoconnect VPN option is turned on.
05-23-2013 10:58 AM
05-23-2013 11:07 AM
05-26-2013 11:21 AM
Mine does it as well....
Z10 STL100-3 with the updated Os of 10.1.0.273 on the Rogers network Canada,
Scenario's are: In bedtime mode with charger plugged in got woken up
at 5.30 am when gmail sent notification to my phone.... Phone rebooted
and bedtime was turned off...
Mornings when I disconnect the charger phone reboots itself...
Phone just sitting on desk doing nothing and it reboots itself...
Happens daily, getting quite frustrated. BLACBERRY GET THIS FIXED ASAP!
Crackberry has a pole that the STL100-3 model appears to be getting
the worst of the reboot issues... If its a battery issue recall them and get the phones
fixed before it burn a home down....
06-04-2013 12:42 AM
has anyone got feedback from Blackberry right now? I have the same issue and it's getting pretty worse.
Using 100-3 Z10, O2 Germany Carrier and Version 10.1.0.273
Started with reboots in bedmode and right now I have at least 2 reboots a day.
Blackberry please help with this Issue.
The phone is really good, but you won't get your feet on the ground again, if this problem won't be solved in time.