Re: Re-initializing a list after the control returns to gdb...

[Andrew Cagney <ac131313@redhat.com>]

> I did it by inserting a call to ada_reset_tasks_list() inside
> normal_stop(), like so:
> 
> *************** print_stop_reason (enum inferior_stop_re
> *** 3360,3365 ****
> --- 3362,3369 ----
>   void
>   normal_stop (void)
>   {
> +   ada_reset_tasks_list (NULL);
> + 
>     /* As with the notification of thread events, we want to delay
>        notifying the user that we've switched thread context until
>        the inferior actually stops.
> 
> My testing shows that it seems to be working pretty well. But do you
> think that this is a good idea? And does this change have any chance of
> being accepted for inclusion (eventually, we are still ironing out some
> wrinkles here and there to make our changes more acceptable, but hope to
> be able to contribute our ada-language support changes soon)?

You've the `observer' problem.

GDB's observer code is very primative.  A combination of gdb-events 
(needs direction) and chained hooks (see target_new_objfile_hook).

I suspect what you really want is a `target changed' event.

Andrew