From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1M8tS9-00033I-F1
	for qemu-devel@nongnu.org; Tue, 26 May 2009 06:00:13 -0400
Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43)
	id 1M8tS3-000323-TA
	for qemu-devel@nongnu.org; Tue, 26 May 2009 06:00:12 -0400
Received: from [199.232.76.173] (port=53857 helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1M8tS3-00031r-FO
	for qemu-devel@nongnu.org; Tue, 26 May 2009 06:00:07 -0400
Received: from smtp02.citrix.com ([66.165.176.63]:59788)
	by monty-python.gnu.org with esmtp (Exim 4.60)
	(envelope-from <Stefano.Stabellini@eu.citrix.com>)
	id 1M8tS3-0006Kt-2G
	for qemu-devel@nongnu.org; Tue, 26 May 2009 06:00:07 -0400
Message-ID: <4A1BBE31.9090106@eu.citrix.com>
Date: Tue, 26 May 2009 11:02:25 +0100
From: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
MIME-Version: 1.0
Subject: Re: [Qemu-devel] Do we need CONFIG_AIO?
References: <20090525082151.GA4107@lst.de> <4A1A85B4.3060200@eu.citrix.com>
	<4A1BA111.9060801@codemonkey.ws>
In-Reply-To: <4A1BA111.9060801@codemonkey.ws>
Content-Type: text/plain; charset="UTF-8"
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: Anthony Liguori <anthony@codemonkey.ws>
Cc: Christoph Hellwig <hch@lst.de>, "qemu-devel@nongnu.org" <qemu-devel@nongnu.org>

Anthony Liguori wrote:

> Stefano Stabellini wrote:
>> Christoph Hellwig wrote:
>>
>>   
>>> Currently all aio code in the raw-posix block driver depends on a
>>> CONFIG_AIO symbol that gets set based on the availability of the
>>> pthreads library.  Beeing able to rely on unconditional aio support
>>> means we could remove the whole code for plain bdrv_read/bdrv_write
>>> in raw-posix.c which is about 300 lines of code and a duplicated code
>>> path.  Is there any Posix-ish platform qemu supports that does not
>>> have pthreads support?
>>>
>>>     
>> MiniOS.
>> I would appreciate if you keep that flag.
>>   
> 
> Threads are not going to be able to remain an optional dependency 
> forever.  I would be willing to merge something to remove CONFIG_AIO 
> although I expect that there will be some fall out.  Some people are 
> using --disable-aio today for either historic reasons or because there 
> are weird bugs with AIO enabled.
> 

Obviously I would prefer if threads remained optional too, but let's
just focus on AIO for the moment: as Todd pointed out OpenBSD doesn't
currently provide AIO and also some small libc, like newlib, do not have
AIO either.