Finalze behavior for uncaught panics in child threads

[aturon]

Currently, an uncaught panic on a child thread will result in output to that thread's panic output handle (which is currently overridable).

This may not be a desirable default, and we probably want the behavior to be more thoroughly customizable.

See this previous RFC for some discussion on the topic, and the corresponding RFC issue.

We need to make a decision here before 1.0.

[aturon]

triage: P-backcompat-lang (1.0 beta)

[pnkfelix]

This is not going to change in time for beta.

In any case, the current behavior can be a default going-forward (it is a default we can live with), and then we can backwards compatibly add the ability to customize the behavior.

Reclassifying: not 1.0 beta, and not 1.0; P-high.

[theemathas]

Related to #23172

[alexcrichton]

With a stable release, the introduction of panic handlers, and the introduction of panic propagation, I'm going to close this issue now.