Feat: A mechanism to set activeDeadlineSeconds

[dalehamel]

We don't want to allow tracejobs to accumulate without bound or accidentally left running, sapping system resources for traces no one will look at.

To achieve this, I think that setting activeDeadlineSeconds on the trace job would provide a built-in way to ensure that no job runs longer than a sane default (say, 1 hour), unless specifically requested with a command-line flag to override the deadline (eg, --deadline 24h) to extend the time that a job is allowed to run.

Ideally, this would also signal bpftrace to dump its map / summarize the map data before exiting, and could also be used as a mechanism for having jobs run for a specific duration. Eg, sample for 10 minutes than exit.