From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:45219)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <famz@redhat.com>) id 1aIxN6-0000nF-JG
	for qemu-devel@nongnu.org; Tue, 12 Jan 2016 06:44:21 -0500
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <famz@redhat.com>) id 1aIxN5-00045m-P7
	for qemu-devel@nongnu.org; Tue, 12 Jan 2016 06:44:20 -0500
Date: Tue, 12 Jan 2016 19:44:13 +0800
From: Fam Zheng <famz@redhat.com>
Message-ID: <20160112114413.GA3903@ad.usersys.redhat.com>
References: <1452483373-22471-1-git-send-email-famz@redhat.com>
	<1452483373-22471-3-git-send-email-famz@redhat.com>
	<5693B59A.6040806@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <5693B59A.6040806@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v2 2/2] nbd: Coroutine based
	nbd_send_negotiate
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: Kevin Wolf <kwolf@redhat.com>, qemu-devel@nongnu.org, qemu-block@nongnu.org

On Mon, 01/11 15:00, Paolo Bonzini wrote:
> 
> 
> On 11/01/2016 04:36, Fam Zheng wrote:
> >  
> > +    aio_set_fd_handler(ctx, client->sock, true,
> > +                       nbd_negotiate_continue,
> > +                       nbd_negotiate_continue, data->co);
> >      TRACE("Beginning negotiation.");
> >      memset(buf, 0, sizeof(buf));
> 
> This causes a busy loop if the socket is writable but the client does
> not send data.  I think you need to set/clear the handler (using
> qemu_coroutine_self() instead of data->co, probably) every time the
> direction of negotiation switches.  That is, set only a read handler
> before read_sync, and only a write handler before write_sync.
> 

You are right, I'll take a look and work on v3.

Fam