From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1Mw0t9-0004at-O2
	for qemu-devel@nongnu.org; Thu, 08 Oct 2009 17:51:07 -0400
Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43)
	id 1Mw0t3-0004UD-Kd
	for qemu-devel@nongnu.org; Thu, 08 Oct 2009 17:51:06 -0400
Received: from [199.232.76.173] (port=54151 helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1Mw0t3-0004Tv-Ca
	for qemu-devel@nongnu.org; Thu, 08 Oct 2009 17:51:01 -0400
Received: from mail-qy0-f199.google.com ([209.85.221.199]:42339)
	by monty-python.gnu.org with esmtp (Exim 4.60)
	(envelope-from <anthony@codemonkey.ws>) id 1Mw0t3-0004pk-1K
	for qemu-devel@nongnu.org; Thu, 08 Oct 2009 17:51:01 -0400
Received: by qyk37 with SMTP id 37so697601qyk.18
	for <qemu-devel@nongnu.org>; Thu, 08 Oct 2009 14:51:00 -0700 (PDT)
Message-ID: <4ACE5EBF.50500@codemonkey.ws>
Date: Thu, 08 Oct 2009 16:50:55 -0500
From: Anthony Liguori <anthony@codemonkey.ws>
MIME-Version: 1.0
Subject: Re: [Qemu-devel] Re: [PATCH] Revert "posix-aio-compat: avoid signal
	race when spawning a thread"
References: <20091007132006.GA9668@redhat.com>	<Pine.LNX.4.64.0910072000090.2178@linmac.oyster.ru>	<4ACCC1A5.5020306@codemonkey.ws>
	<20091007204421.GC4085@redhat.com>	<Pine.LNX.4.64.0910080053350.4678@linmac.oyster.ru>	<20091007205523.GA4783@redhat.com>	<Pine.LNX.4.64.0910080059550.4776@linmac.oyster.ru>	<20091008014742.GA15601@shareable.org>	<Pine.LNX.4.64.0910080647510.2109@linmac.oyster.ru>	<20091008151527.GB29691@shareable.org>
	<20091008212421.GA7229@redhat.com>
In-Reply-To: <20091008212421.GA7229@redhat.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
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: "Michael S. Tsirkin" <mst@redhat.com>
Cc: qemu-devel@nongnu.org

Michael S. Tsirkin wrote:
> BTW, tried looking at the code.
> I saw some (unrelated) issues:
>
> static void aio_signal_handler(int signum)
> {
>     if (posix_aio_state) {
>         char byte = 0;
>
>         write(posix_aio_state->wfd, &byte, sizeof(byte));
>     }
>
>     qemu_service_io();
> }
>
> And qemu_service_io does a *ton* of things.
>   

It just does a cpu_exit().

> Questions:
> -  Do we need the call to qemu_service_io? Seems to
>    behave the same with and without it.
>   

Where you run into trouble is when you have a guest that has no periodic 
timer  and needs a mechanism to break out of the tcg loop whenever an IO 
operation completes.  See


Author: aliguori <aliguori@c046a42c-6fe2-441c-8c8c-71466251a162>
Date:   Wed Oct 8 19:50:24 2008 +0000

    Fix IO performance regression in sparc

> -  Are all of the data structures touched by qemu_service_io
>    protected by blocking signals before access?
>   

cpu_exit is pretty careful to not do anything that could be racey.

> Also:
> 	- let's use signalfd on linux, if available?
>   

signalfd doesn't help.  In fact, we originally used signalfd and it 
caused a regression.

> 	- for SIGALARM, maybe timerfd?
>   

Or just get rid of SIGALARM...

Regards,

Anthony Liguori