Reboot - Diagnostics?


#1

We developed an application running on a Fastrack Supreme 10 / Open AT OS. After some recent changes the application developed a habit of rebooting the device. This only happens under certain conditions and is most likely the cause of some memory leak.

I added trace statements to identify where the application is rebooting but all of the user code executes successful. To my best knowledge it appears to crash/reboot while waiting for a timer to elapse. (It does not appear to happen just before the next elapse either, but rather a couple of seconds after it finished the last one. Commenting out all the code in the function executed befer does not change this behaviour either.)

[]Are there any diagnostic tools I can use to further investigate this behaviour?[/]

[]Any ideas what could cause behaviour like that?[/]

Somewhere on this forum I read that not explicitly casting pointer variables before doing a wm_memcpy() might cause a reboot. I will double check that, but have little hope this will fix the issue as it does not happen at a wm_memcpy.

Thanks for any advice what so ever!
Christian


#2

If you are waiting for a timer, have you allowed time for the task to switch to the watchdog kicker? Tasks must be released regularly to allow the watchdog task to run.
Is there any mention of the watchdog task in the diagnosics read out via TMT app?


#3

You mean “caused by some memory leak”, I presume…?

What do you mean by that?

You’re not using a busy loop to “wait”, are you…?

See FAQ:
viewtopic.php?f=7&t=3766#p16098
viewtopic.php?f=7&t=3766&start=15#p18220