Comment 7 for bug 1183065

A couple of comments from the log:

* powerd tries to suspend the device, but often suspend is being blocked by smd_sns_dsps. I suspect this has something to do with sensors. Note that on the nexus 4 the driver for the sensors is completely in userspace, so it isn't tied into the kernel suspend/resume mechanism. We may need to get the sensor code listening for powerd state change notifications to quiesce the hardware before we suspend.

* There are a lot of wakeups from suspend. The logs don't always tell us the source of the wakeup, but there are a lot of "wakeup wake lock: smd_sns_dsps" messages which tell that this lock is taken very shortly after resuming. Also note the "suspend: too many immediate wakeups, back off" messages, which indicate that the device is frequently going into suspends for very short durations of time.