Hi Dave,
On 10/10/2012 11:35 AM, Dave Pigott wrote:
Hi all,
I found an interesting health failure today on origen07
http://validation.linaro.org/lava-server/scheduler/job/35016/log_file
When you look at the log, you see that the board starts off at the u-boot prompt. It then tries to do a "reboot", which (obviously) fails. So naturally, it then does a hard reset, and this is where it does something very odd: It interrupts the boot and tries to boot the previously installed test image. I haven't yet looked at the dispatcher code to figure out why (that's my next job).
What then started alarm bells ringing was that I saw this:
1261680 bytes read reading uInitrd
1532597 bytes read reading board.dtb
** Unable to read "board.dtb" from mmc 0:5 **
So whatever the test image was, it was expecting a device tree blob, which I would have assumed would have to have been installed during deploy_linaro_image() being that if there is one it should just be part of the test boot deployment.
So I looked at the log from the previous job:
http://validation.linaro.org/lava-server/scheduler/job/34938/log_file#entry2...
and sure enough, you'll see at that mark the same issue.
So there are two things:
- There's some twisted logic in the dispatcher that's making it do odd things if it starts off in u-boot
- Do we have an issue with dtbs not being handled properly by lava, or is it just that the hwpack was incomplete?
Regarding the 2).
No dtb in the hwpack. I've created a very similar bug for another ci project: https://bugs.launchpad.net/linaro-ci/+bug/1064686
Probably "ubuntu packed kernels" is the only project (view) in jenkins handling the dtb in the right way.
Please note, that the dtb must be properly described in the hwpack, so that it could be picked up by LAVA / l-m-c. This is out of scope of the bug 1064686.
Thanks, Andrey