From owner-freebsd-hackers  Fri Apr 26 15:15:56 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id PAA08009
          for hackers-outgoing; Fri, 26 Apr 1996 15:15:56 -0700 (PDT)
Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id PAA08004
          for <hackers@FreeBSD.org>; Fri, 26 Apr 1996 15:15:53 -0700 (PDT)
Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id PAA28041; Fri, 26 Apr 1996 15:10:00 -0700
From: Terry Lambert <terry@lambert.org>
Message-Id: <199604262210.PAA28041@phaeton.artisoft.com>
Subject: Re: Something fishy with our PT_ATTACH code!
To: jkh@time.cdrom.com (Jordan K. Hubbard)
Date: Fri, 26 Apr 1996 15:10:00 -0700 (MST)
Cc: sef@kithrup.com, hackers@FreeBSD.org
In-Reply-To: <229.830534797@time.cdrom.com> from "Jordan K. Hubbard" at Apr 26, 96 09:06:37 am
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-hackers@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

> > It doesn't seem that the detached process is suspending itself so much as it
> > is simply renotifying the parent that it is still stopped.  This could be
> > wrong behaviour, however.
> > 
> > To change that, in the PT_DETACH code, a
> > 
> > 	psignal(p, SIGCONT);
> > 
> > would do the trick.  I think ;).
> 
> Well, it seems to me that we should really restore the process's
> previous state - what if it was suspended to start with? :-)


begin:
	initial_state

attach_1:
	current_state	->	save_state_1	(initial_state)
	suspend_state	->	current_state

attach_2:
	current_state	->	save_state_2	(attach_1_suspend)
	suspend_state	->	current_state

detach_1:
	save_state_1	->	current_state	(initial_state)

detach_2:
	save_state_2	->	current_state	(attach_1_syspend)

end:
	attach_1_suspend


You need to modify attack_2's save_state_2 based on entry order
when you detach_1.  You need an attach list back-pointer.


Or you need to imply O_EXCL and prevent reeentrancy.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.