maliit-server health-check failures

[Paul Larson]

We ran baselines of many processes on an idle nexus4 using health-check for the last trusty image, and some of those have regressed since then. These could be real regressions, or they could be something that's expected due to changes that happened since that image. We need someone to investigate these and report whether it requires a fix, or whether the thresholds need to be adjusted, and why.

Results from a recent test run can be seen at http://ci.ubuntu.com/smokeng/utopic/touch/mako/204:20140822:20140811.1/9853/health-check/

There were some pretty large changes in maliit-server - CPU went from ~.01 to ~2.0 for system and overall cpu usage! An otherwise idle process eating 2% of the cpu seems a little odd to me. It also looks like wakeups went up somewhat, which we need to either understand the reason for, or raise the threshold on them.