cvs commit: src/sys/kern kern_conf.c

[Tor Egge]

> This is quite likely to not happen without intervention.  It might
> be a good idea to to deliver a signal to the process, if we can find it.

The process might have been woken but not yet scheduled due to the caller of
destroy_dev() having called something similar to the purge method before
calling destroy_dev().

> Better of course would be more purge methods.

Agreed.  Pseudo ttys currently lack a purge method.

- Tor Egge