# How can I get a backtrace of device crashes when connecting to mobile Internet? (1.0.8.19)) [duplicate]

I notice frequent crashes of the device. It just turns off. Many times, after the restart it crashes right again.

I assume this is related to mobile internet access for a couple of reasons:

• I never saw these crashes after skipping the PIN-code. i.e. without GSM access
• I never saw these crashes in flight-mode
• It seems to crash more likely when in areas with poor network connection
• Crashes often happen as soon as browser starts

Is there any way I can provide a backtrace or any other further information that will help you fixing the issue?

EDIT: Another strange thing I noticed is that after such a crash, the device often claims that the battery had only 4% voltage left. After another reboot it indicates a plausible value (e.g. 50%) again.

edit retag reopen delete

### The question has been closed for the following reason "duplicate question"by nthn close date 2014-10-11 01:14:22.438752

I think this is a duplicate of

https://together.jolla.com/question/7144/jolla-randomly-shuts-down-10819/

Even if the title only mentions reboots, much people have the problem, that the device just turns off.

( 2014-09-26 21:59:41 +0200 )edit

Yes, I saw this too and it looks similar indeed. I don't think that my crashes are because of the battery, though, because my device never crashed when then SIM-card was inactive. If the connection to the battery wasn't sufficient, the device would crash regardless of the SIM, wouldn't it?

( 2014-09-27 02:03:04 +0200 )edit

Sort by » oldest newest most voted

I applied the "fix" using a piece of paper as suggested in https://together.jolla.com/question/7144/jolla-randomly-shuts-down-10819/. I didn't expect any improvement but to my surprise there were no crashes anymore in more than a week.

Thus, this report here seem to be a duplicate of https://together.jolla.com/question/7144/jolla-randomly-shuts-down-10819/, indeed. It can be closed. I'll report my findings there.

more

Together with the battery issue, it looks more like a hardware problem (some part of the phone draws too much current when it's activated, maybe. Possibly a bad solder point somewhere, or a crossconnection). If this is the case, logs won't help very much, and it will be probably difficult to obtain logs in any case.

Have you contacted the Jolla support?

Anyway, here's what I would try to get some logs. Activate developer mode, and connect to the phone via ssh over usb. Use devel-su to switch to root, then journalctl -f to follow the syslog. Then try to reproduce the crash. If you're lucky, you might see something before the phone dies, but I doubt it.

There are various other logs for the other cores (the GSM modem runs on a pair of different DSP cores) and various bits of other hardware devices, but I haven't completely figured out all of that. The logs won't be much use anyway, because we don't know how these parts of the phone work, and probably nobody outside of Qualcomm does.

Contacting support and describing how to reproduce the problem is probably your best option in my opinion.

more

No, I did not contact Jolla yet (except here).

Just to be clear: The battery is not actually empty after the reboot. The phone reports a low voltage (reproducibly 4%), but after another reboot it is full again.

I give the "fix" with the piece of paper at the battery a try which is described at the possible duplicate. I'll let you know how that works...

( 2014-09-29 23:45:34 +0200 )edit

@RobertLoehning: Yes, that's how I understood it. It would be consistent with the battery being taxed heavily by a strong current, not actually discharging the battery but putting it in a state where it reports low voltage until the battery (or the control circuitry) recovers. Using a piece of paper only helps for bad contact between battery and phone causing intermittent disconnections and power failure, and won't help in case of a hardware problem causing a strong current somewhere else in the phone. Of course, it's all guesswork so far.

( 2014-09-30 12:16:03 +0200 )edit