RFC: Targets that do not call generic_mourn_inferior() ?

[Andrew Cagney]

Fernando Nasser wrote:
> 
> It seems to me that, except for very exceptional circustances, all
> remote targets should be calling generic_mourn_inferior().  It does some
> gdb cleanup that is necessary if the user ever reconnects to the target
> within the same session.
> 
> An alternative is to have gdb forcifully calling this when unpushing a
> target, but the current way is more flexible (and more error prone as
> well).
> 
> Here is a patch for two targets that missed that call and act very silly
> if one does a "I reg" or "bt" after reconnecting.

Approved (and checked in).

	Andrew