Android phone reboots while executing my own developed application

Question

I developed an Android application that uses many services, including GPS and orientation (compass). The genre of the application is AR (augmented reality) and I'm moving image views around the screen in real-time depending on the phone orientation. However, after a while, the phone freezes for a second and then reboots (Motorola Droid). I'm unable to find any useful log about this event. The log in the Eclipse is cleared after the phone reboots; I use "adb logcat" command to write the log to a file on my PC. The last meaningful rows of information are:

I/ActivityManager( 1272): Displayed activity com.zlatko/.Main: 2656 ms (total 2656 ms)

D/CameraHal( 1050): Zoom callback param1: 65536, param2: 1

D/CameraHal( 1050): processZoom/1:0. stopped=1

D/CameraHal( 1050): Stored Zoom Data/0/1:0

D/dalvikvm( 1272): GC freed 8444 objects / 453552 bytes in 268ms

D/dalvikvm( 1656): GC freed 13158 objects / 456056 bytes in 257ms

D/dalvikvm( 1656): GC freed 23844 objects / 787840 bytes in 98ms

D/dalvikvm( 1656): GC freed 22942 objects / 756408 bytes in 94ms

How can I find the reason for this frustrating behavior? The logs don't offer insight, there are no explicit exceptions thrown, the system is indeterminate (that means, I cannot conclude at what point in time exactly the phone reboots)...

Any recommendation?

Thanks, Zlatko

Answer 1

The bugreport you linked to has a possible answer in the "KERNEL PANIC CONSOLE LOG" section:

<1>[ 1845.425445] Unable to handle kernel NULL pointer dereference at virtual address 00000000
<1>[ 1845.425811] pgd = cdfd4000
<1>[ 1845.426025] [00000000] *pgd=8d16d031, *pte=00000000, *ppte=00000000
<4>[ 1845.426879] Internal error: Oops: 817 [#1] PREEMPT
<4>[ 1845.427185] Modules linked in: tiwlan_drv sec modem_pm_driver netmux_linkdriver netmux
<4>[ 1845.428283] CPU: 0    Not tainted  (2.6.29-omap1 #1)
<4>[ 1845.428497] PC is at SGXGetMiscInfoKM+0x310/0x348
<4>[ 1845.428680] LR is at List_PVRSRV_DEVICE_NODE_ForEach+0x1c/0x2c

Looks similar to this bug report for a Milestone and a post from May 22 2010 on this Motorola support forum.