From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1MbqRt-0006Rr-9z
	for qemu-devel@nongnu.org; Fri, 14 Aug 2009 02:39:37 -0400
Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43)
	id 1MbqRo-0006RM-Lp
	for qemu-devel@nongnu.org; Fri, 14 Aug 2009 02:39:36 -0400
Received: from [199.232.76.173] (port=46384 helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1MbqRo-0006RJ-EM
	for qemu-devel@nongnu.org; Fri, 14 Aug 2009 02:39:32 -0400
Received: from mx2.redhat.com ([66.187.237.31]:58035)
	by monty-python.gnu.org with esmtp (Exim 4.60)
	(envelope-from <markmc@redhat.com>) id 1MbqRn-0004Jo-R1
	for qemu-devel@nongnu.org; Fri, 14 Aug 2009 02:39:32 -0400
Subject: Re: [Qemu-devel] [PATCH 2/5] Add SCM_RIGHTS support to unix socket
	character devices
From: Mark McLoughlin <markmc@redhat.com>
In-Reply-To: <4A843D4A.5030306@cs.ualberta.ca>
References: <1248250235.2867.34.camel@blaa>
	<1248250302-28595-2-git-send-email-markmc@redhat.com>
	<4A843D4A.5030306@cs.ualberta.ca>
Content-Type: text/plain
Date: Fri, 14 Aug 2009 07:38:39 +0100
Message-Id: <1250231919.3010.9.camel@blaa>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Reply-To: Mark McLoughlin <markmc@redhat.com>
List-Id: qemu-devel.nongnu.org
List-Unsubscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/pipermail/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Cam Macdonell <cam@cs.ualberta.ca>
Cc: qemu-devel@nongnu.org

On Thu, 2009-08-13 at 10:20 -0600, Cam Macdonell wrote:
> > @@ -1957,6 +2004,10 @@ static void tcp_chr_read(void *opaque)
> >              tcp_chr_process_IAC_bytes(chr, s, buf, &size);
> >          if (size > 0)
> >              qemu_chr_read(chr, buf, size);
> > +        if (s->msgfd != -1) {
> > +            close(s->msgfd);
> > +            s->msgfd = -1;
> > +        }
> 
> Hi Mark,
> 
> I'm trying to understand the intended behaviour here so I can use your 
> patch.  In your comments for the patch it reads "store the file 
> descriptor for use in the chr_read() handler.  Close the file descriptor 
> if it was not used."  It seems that upon returning from the chr_read() 
> handler the descriptor is closed if s->msgfd is not set to -1.  Does 
> this mean the handler should set the fd to -1 after retrieving it if the 
> process wants to keep that fd open, otherwise it will be closed when 
> chr_read() returns?  Should tcp_get_msgfd() perhaps set  s->msgfd to -1 
> after it is retrieved?  Sort of a "read and clear" behaviour?

Yes, that would work and TBH I can't recall why I chose not to do that.
It'd be like popping the fd from a one item queue.

The current caller of get_msgfd() dup()s the fd which achieves the same
thing.

Cheers,
Mark.