From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:33294)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1VCRYX-0007Bh-23
	for qemu-devel@nongnu.org; Thu, 22 Aug 2013 05:51:59 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1VCRYR-0007sp-1x
	for qemu-devel@nongnu.org; Thu, 22 Aug 2013 05:51:52 -0400
Received: from mx1.redhat.com ([209.132.183.28]:34317)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1VCRYQ-0007sg-PE
	for qemu-devel@nongnu.org; Thu, 22 Aug 2013 05:51:46 -0400
Message-ID: <5215DF04.3040601@redhat.com>
Date: Thu, 22 Aug 2013 11:51:00 +0200
From: Paolo Bonzini <pbonzini@redhat.com>
MIME-Version: 1.0
References: <1377050567-19122-1-git-send-email-asias@redhat.com>
	<52147742.2090401@redhat.com>
	<20130822095022.GA25895@hj.localdomain>
In-Reply-To: <20130822095022.GA25895@hj.localdomain>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH] block: Fix race in gluster_finish_aiocb
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: Asias He <asias@redhat.com>
Cc: Kevin Wolf <kwolf@redhat.com>, Vijay Bellur <vbellur@redhat.com>, qemu-devel@nongnu.org, Stefan Hajnoczi <stefanha@redhat.com>, Bharata B Rao <bharata@linux.vnet.ibm.com>, MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>

Il 22/08/2013 11:50, Asias He ha scritto:
> On Wed, Aug 21, 2013 at 10:16:02AM +0200, Paolo Bonzini wrote:
>> Il 21/08/2013 04:02, Asias He ha scritto:
>>> In block/gluster.c, we have
>>>
>>> gluster_finish_aiocb
>>> {
>>>    if (retval != sizeof(acb)) {
>>>       qemu_mutex_lock_iothread(); /* We are in gluster thread context */
>>>       ...
>>>       qemu_mutex_unlock_iothread();
>>>    }
>>> }
>>>
>>> qemu tools, e.g. qemu-img, might race here because
>>> qemu_mutex_{lock,unlock}_iothread are a nop operation and
>>> gluster_finish_aiocb is in the gluster thread context.
>>>
>>> To fix, we introduce our own mutex for qemu tools.
>>>
>>> Signed-off-by: Asias He <asias@redhat.com>
>>> ---
>>>  stubs/iothread-lock.c | 11 +++++++++++
>>>  1 file changed, 11 insertions(+)
>>>
>>> diff --git a/stubs/iothread-lock.c b/stubs/iothread-lock.c
>>> index 5d8aca1..d5c6dec 100644
>>> --- a/stubs/iothread-lock.c
>>> +++ b/stubs/iothread-lock.c
>>> @@ -1,10 +1,21 @@
>>>  #include "qemu-common.h"
>>>  #include "qemu/main-loop.h"
>>>  
>>> +static QemuMutex qemu_tools_mutex;
>>> +static pthread_once_t qemu_tools_once = PTHREAD_ONCE_INIT;
>> 
>> Doesn't work on Windows, but you can just add
> 
> Hmm, Any reasons, why it does not work on Windows?

There are no pthreads on Windows.

There is an emulation library, but we're not using it.

>> __attribute__((__constructor__)) to qemu_tools_mutex_init.
> 
> __attribute__((__constructor__)) works on Windows?

Yes, it is part of the runtime library's support for C++.  We use it
already, see include/qemu/module.h.

Paolo

>> Paolo
>>
>>> +static void qemu_tools_mutex_init(void)
>>> +{
>>> +    qemu_mutex_init(&qemu_tools_mutex);
>>> +}
>>> +
>>>  void qemu_mutex_lock_iothread(void)
>>>  {
>>> +    pthread_once(&qemu_tools_once, qemu_tools_mutex_init);
>>> +    qemu_mutex_lock(&qemu_tools_mutex);
>>>  }
>>>  
>>>  void qemu_mutex_unlock_iothread(void)
>>>  {
>>> +    qemu_mutex_unlock(&qemu_tools_mutex);
>>>  }
>>>
>>
>