From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:52543)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <famz@redhat.com>) id 1ZYoww-0007zW-Gq
	for qemu-devel@nongnu.org; Mon, 07 Sep 2015 01:26:39 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <famz@redhat.com>) id 1ZYowv-0004SM-Ov
	for qemu-devel@nongnu.org; Mon, 07 Sep 2015 01:26:38 -0400
Date: Mon, 7 Sep 2015 13:26:29 +0800
From: Fam Zheng <famz@redhat.com>
Message-ID: <20150907052629.GB3958@ad.nay.redhat.com>
References: <1438144934-23619-1-git-send-email-famz@redhat.com>
	<1438144934-23619-8-git-send-email-famz@redhat.com>
	<20150827172338.GG4917@stefanha-thinkpad.redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20150827172338.GG4917@stefanha-thinkpad.redhat.com>
Subject: Re: [Qemu-devel] [PATCH v2 07/11] aio-posix: introduce aio_{disable,
 enable}_clients
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: Stefan Hajnoczi <stefanha@gmail.com>
Cc: kwolf@redhat.com, pbonzini@redhat.com, stefanha@redhat.com, qemu-devel@nongnu.org, qemu-block@nongnu.org

On Thu, 08/27 18:23, Stefan Hajnoczi wrote:
> On Wed, Jul 29, 2015 at 12:42:10PM +0800, Fam Zheng wrote:
> > +void aio_disable_enable_clients(AioContext *ctx, int clients_mask,
> > +                                bool is_disable)
> > +{
> > +    AioHandler *node;
> > +    aio_context_acquire(ctx);
> > +
> > +    QLIST_FOREACH(node, &ctx->aio_handlers, node) {
> > +        if (!node->deleted && node->type & clients_mask) {
> > +            node->disable_cnt += is_disable ? 1 : -1;
> > +        }
> > +    }
> > +    aio_context_release(ctx);
> > +}
> 
> If someone adds an fd of a disabled type *after* the call to
> aio_disable_clients() then it won't be disabled.
> 
> Another approach is to keep an array of counters per AioContext and
> check the counters during aio_poll() when deciding which fds to monitor.

Good idea, I'll change that way.

> 
> Also, this function acquires/releases AioContext so it's worth
> mentioning in the doc comments that this function is thread-safe.
> 

OK.

Fam