From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([140.186.70.92]:50576)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <paolo.bonzini@gmail.com>) id 1R1z0i-0000Qp-1e
	for qemu-devel@nongnu.org; Fri, 09 Sep 2011 07:12:44 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <paolo.bonzini@gmail.com>) id 1R1z0h-0004QG-1m
	for qemu-devel@nongnu.org; Fri, 09 Sep 2011 07:12:40 -0400
Received: from mail-vx0-f173.google.com ([209.85.220.173]:40375)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <paolo.bonzini@gmail.com>) id 1R1z0g-0004Q6-VW
	for qemu-devel@nongnu.org; Fri, 09 Sep 2011 07:12:39 -0400
Received: by vxj15 with SMTP id 15so666342vxj.4
	for <qemu-devel@nongnu.org>; Fri, 09 Sep 2011 04:12:38 -0700 (PDT)
Sender: Paolo Bonzini <paolo.bonzini@gmail.com>
Message-ID: <4E69F4A2.3070405@redhat.com>
Date: Fri, 09 Sep 2011 13:12:34 +0200
From: Paolo Bonzini <pbonzini@redhat.com>
MIME-Version: 1.0
References: <1315561022-25386-1-git-send-email-stefanha@linux.vnet.ibm.com>
In-Reply-To: <1315561022-25386-1-git-send-email-stefanha@linux.vnet.ibm.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH 0/2] Make simpletrace work on Windows
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@linux.vnet.ibm.com>
Cc: Jan Kiszka <jan.kiszka@siemens.com>, qemu-devel@nongnu.org

On 09/09/2011 11:37 AM, Stefan Hajnoczi wrote:
> The 'simple' trace backend uses pthreads and does not work on Windows.  These
> patches switch from pthreads to glib so that the code builds on all platforms
> supported by glib.
>
> Only one thing I'm unhappy about: the simpletrace write-out thread used to
> block all signals.  I have removed that code and don't expect glib to do it for
> me.  I'm not sure if there is a problem if signal handlers are invoked in the
> write-out thread instead of a QEMU thread.  Any thoughts?

I am afraid of concurrent manipulation of the alarm timer info (in 
host_alarm_handler) if SIGALRM is triggered in the write-out thread. 
This would be with -clock unix.  I'd just leave the pthread_sigmask code 
#ifdef'ed in.

By the way, I think this is the exact problem that Windows has running 
with the dynticks timer instead of the mmsystem timer.  We can fix this 
particular problem by transferring host_alarm_handler's work to a bottom 
half, but in general I think leaving signals enabled is dangerous.

Paolo