Re: [RFC]: Remote Protocol "attach" command.

[Andrew Cagney <ac131313 at cygnus dot com>]

"J.T. Conklin" wrote:

> I didn't realize that the existing detach packet was used like that.
> I 100% agree that no packet should effect the transport layer.

Yep, the name/doco is really misleading.  I've been given a chance to
work a little on remote.c so expect some further doco changes (finally).

> However, if there is a way to attach to the target (where attach is
> distinct from establishing a connection to it), I believe that there
> should also be some way to detach from it and have it continue as if
> nothing happened.

At present there isn't a way to do an attatch - hence Michael's e-mail. 
I think there are several operations required:

	o	select the program to run

	o	restart the program

	o	attatch to a running program

	o	detatch from a running program

For those that are curious, this all goes back to ``target remote'' vs
``target extended-remote''.  The former assumes that the remote stub
quits on program exit while the former assumes that the remote stub is
long running (ie runs for longer than the program being debugged).

	Andrew