Hi All,
I am trying to rule out SW issues with my BBB (HW Rev A5C) running Angstrom. First I am going to give a bit of background on when the issues started and then provide dumps of crashed as seen via UART.
-
BBB was working fine for quite some time.
-
Then I tried to switch from wired ethernet to wireless by plugging in UWN200 (http://www.logicsupply.com/components/networking/wireless/uwn200/) while BBB was powered by DC adapter.
-
My SSH over wired Ethernet froze.
-
I removed the UWN200 and rebooted a device but could not longer connect via SSH with wired ethernet
-
Connecting via USB (client) I could not get network connection to 192.168.7.2 either. As I realized later I did not even have usb0 listed when running ifconfig
-
At this point I connected via UART
-
I could see that the kernel was crashing with NULL pointer dereference Oops
-
Every few restarts it would get past this oops
-
I shutdown BBB and looked to upgrade to latest Angstrom a few days later, thinking maybe I am just a lucky one to hit a real bug in the kernel
-
I downloaded the following image https://s3.amazonaws.com/angstrom/demo/beaglebone/BBB-eMMC-flasher-2013.09.04.img.xz into SD card
-
BBB was booted from SD by holding the boot button.
-
Flashing process was problematic, it took a few reboots (about) from SD card for eMMC flashing to complete (4 solid LEDs) without kernel Oopsing
- One of the Oopses was odd as it was for “invalid instruction”
To me this feels like a HW problem due to failures manifesting themselves in different ways. Segfaults in the apps, NULL pointer dereferences in the kernel, Soft CPU lockups. But I would like to rule out a user error.
Few observations:
-
I have run mtest from uboot without any errors being detected
-
Same for memtester
-
Pressing rest button does not always reboot the device. Sometimes holding power button down is the only way to reboot the BBB. But sometimes it works just fine.
-
For a few reboots accessing BBB via HTTP would causes kernel Oopses. Trying it out just now and it worked just fine
The below is done post flashing to the latest angstrom:
- systemd once segfaulted during boot: http://pastebin.com/YFuxNs8e
- some times system boots just fine: http://pastebin.com/7HynL8H7
- “opkg upgrade” consistently segfaults
- “opkg install strace” consistently segfaults
Any suggestions on what I can do to narrow down the root cause of the issue would be greatly appreciated.
Thank you,
Vladimir