reboot VM after upgrade if kernel or boot scripts got modified

[Martin Pitt]

For tests like LXC we want to run them against the -proposed kernel, so that we can hold back kernel updates which break LXC in -proposed. But for that we need to run the test against the -proposed kernel.

Right now we only dist-upgrade which still runs the kernel in release. We should check in the dist-upgrade if the kernel got updated and if so reboot the VM before continuing.

[Launchpad Janitor (janitor)]

This bug was fixed in the package autopkgtest - 2.17.1

---------------
autopkgtest (2.17.1) unstable; urgency=medium

  * Make $ADTTMP readable and accessible to other users, for tests which use
    other users through "su" and similar.
  * adt-virt-qemu: Set $HOME for root also after rebooting.
  * tests/adt-run: Force stdout to UTF-8 to see proper failures even in
    non-UTF-8 environments, instead of crashing with UnicodeEncodeError.
  * adt-run: Flush input to the stdout/err cat processes and give them some
    time to catch up after the end of the test, to avoid losing the last bits
    of output. (Closes: #750343)
  * Add missing build-essential and debhelper test dependencies.
  * adt-run: Ensure the cleanup handler for the tests tree runs also if
    copying up the tests tree fails. (Closes: #749655)
  * Put back the LXC runner on Debian, so that folks with manual LXC container
    setups can still use it.
  * adt-build-lxc: Check for default networking in /etc/lxc/default.conf. If
    it's "empty", abort with an error message that container networking needs
    to be set up first. Until LXC gets fixed in Debian (#747914) you need to
    manually set up a bridge, dnsmasq, and the default config.
  * adt-build-lxc: Support specifying $ADT_APT_PROXY. If apt-cacher-ng is
    running locally, use that as a default.

 -- Martin Pitt <email address hidden> Wed, 04 Jun 2014 15:03:30 +0200