core: rework how we reset the TTY after use by a service

core: rework how we reset the TTY after use by a service

This makes two changes:

1. Instead of resetting the configured service TTY each time after a
   process exited, let's do so only when the service goes back to "dead"
   state. This should be preferable in case the started processes leave
   background child processes around that still reference the TTY.

2. chmod() and chown() the TTY at the same time. This should make it
   safe to run "systemd-run -p DynamicUser=1 -p StandardInput=tty -p
   TTYPath=/dev/tty8 /bin/bash" without leaving a TTY owned by a dynamic
   user around.