← Back to team overview

ubuntu-phone team mailing list archive

Re: [Canonical-ci-engineering] Mako testing grinding to a halt with image 274

 

Am Donnerstag, den 03.04.2014, 11:28 +0100 schrieb Alan Pope:
> On Thu, Apr 3, 2014 at 10:46 AM, Alan Pope <alan.pope@xxxxxxxxxxxxx> wrote:
> > On Thu, Apr 3, 2014 at 10:28 AM, Oliver Grawert <ogra@xxxxxxxxxx> wrote:
> >> i captured the /proc/last_kmgs via recovery from the device right
> >> after ...
> >> http://paste.ubuntu.com/7197815/
> >> (after 977 seconds i pressed the power button) ...
> >
> > Here's my /proc/last_kmsg, I hard rebooted after a few minutes in activity
> >
> > http://paste.ubuntu.com/7197823/
> 
> That should be http://paste.ubuntu.com/7197970/ sorry.

so adding:

echo "initrd: Running run-init ..." >>/proc/kmsg || true 

to the /init script in the initrd and adding:

 echo "Running mountall ..." >>/proc/kmsg || true

to /etc/init/mountall.conf (mountall is the very first job that upstart
runs after the run-init switch)

This results in the following log at the hang:

[    3.597436] initrd: mounting /root/var/lib/lxc/android/system.img
as /root/android/system
[    3.779185] Adding 524284k swap on /root/userdata/SWAP.img.
Priority:-1 extents:5 across:557052k SS
[    3.955074] initrd: Running run-init ...
[  419.929497] PMIC Initiated shutdown cpu_power_off cpu=0
[  419.929528] PMIC Initiated shutdown cpu_power_off cpu=3
[  419.929528] PMIC Initiated shutdown cpu_power_off cpu=2
[  419.929528] PMIC Initiated shutdown cpu_power_off cpu=1
[  419.930047] Powering off the SoC
[  419.930169] Calling scm to disable arbiter
[  419.930352] SCM returned even when asked to busy loop rc=-4
[  419.930444] waiting on pmic to shut msm down

This clearly shows we are not making it past run-init at all in the case
where the hang occurs, i.e. we are never able to switch to the rootfs
and run /sbin/init 
I am not sure at all what to make out of this, since nothing related to
the kernel or initrd changed since last monday ... so we would have seen
the issue before if it is related to any initrd or kernel changes.

I'd be open for any ideas here ... 
http://people.canonical.com/~ogra/touch-image-stats/274.changes has the
list of changes for this image ...

ciao
	oli

Attachment: signature.asc
Description: This is a digitally signed message part


Follow ups

References