Board inaccessible. Updated with wrong Kernel image

Hi guys

I updated the kernel to this version with

Now I cant access the board at all. No SSH or I cant see the board on my PC via USB.
All four user leds and the power led is lit.
Do I need to flash a new Debian image or is there another recovery technique?
If I flash a new image will all my files be wiped clean?

Thanks in advance…


I just had a similar problem after installing kernel 3.14.31-ti-r49 via apt-get. On serial console the system stopped just showing “Starting kernel…”.
I was able to go back to the previously installed kernel 3.14.29-ti-r46 by changing the “uname_r=” entry in /boot/uEnv.txt



Based on that error, you must have the original May 2014 release
flashed to the eMMC. The *.dtb's in v3.14.x-ti and mainline are too
"fat" for the original memory positions we used in that release. and
corrupt the kernel image on bootup... It hard locks pretty quick..

The new future proof magic number for u-boot are:


1: reflash (using a newer image: )

2: boot with the microSD and edit "uEnv.txt" in the first partition.



Can you your serial boot log with "3.14.31-ti-r49" i
didn't see that issues yesterday with my bbb.


Thanks guys
I really appreciate your help
I flashed the board with a new 3.8.13-bone50 image and all seems to be good.
It worked first time. This guy helped alot too. Thanks Derek

Hi Robert,

I just checked 3.14.31-ti-r49 again and found my problem. I used a custom dtb which was not available in the corresponding directory for the new kernel. Now it’s booting without problems.



To be safe, for the next upgrade, create a directory /boot/dtbs/ and
copy your custom *.dtb into it.

It's the 5th location searched, but first non kernel version directory: