canonical-ci-engineering team mailing list archive

[Paul Larson <paul.larson@xxxxxxxxxxxxx>]

That's good to hear, we've been struggling around the mtp issues for a
while now, and having a cleaner solution would be good. I doubt that mtp is
your problem though. I have never seen the mtp/adb conflict cause a device
to completely go away. It only tends to cause adb disconnects, and
temporary device-not-found at the most.

How is the way we flash our devices in image testing different from how we
flash for these tests? I have a maguro here, and would be interested in
running a continuous flash on it with the same exact procedure that's
causing it to fail for you to see if I can reproduce.  Similarly, maybe it
would be worth adopting our provision.sh script used in image testing? It
has been working reliably for us for quite a while.


On Wed, Nov 6, 2013 at 11:15 AM, Francis Ginther <
francis.ginther@xxxxxxxxxxxxx> wrote:

> On Wed, Nov 6, 2013 at 9:25 AM, Andy Doan <andy.doan@xxxxxxxxxxxxx> wrote:
> > Disabling MTP sounds interesting - However, AIUI, MTP is still going to
> > cause havoc on the target that's really been our core problem. Still
> might
> > be worth a shot given the pain we are currently experiencing.
>
> I looked at disabling MTP connections on the host, turns out it
> already is disabled on okiku, at least I could not find any mtp rules
> under /lib/udev/rules.d/. Those rules were present on
> ps-android-sandybridge. So, not a solution.
>
> For MTP on the target, I learned it can be disabled with:
> "adb shell echo manual > /etc/init/mtp-server-bootup.override"
>
> Sounds like sergio is working on a phablet-* way to do this.
>
> Francis
> --
> Francis Ginther
> Canonical - Ubuntu Engineering - Continuous Integration Team
>
> --
> Mailing list: https://launchpad.net/~canonical-ci-engineering
> Post to     : canonical-ci-engineering@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~canonical-ci-engineering
> More help   : https://help.launchpad.net/ListHelp
>