From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1KyMx2-0004w6-Nd for qemu-devel@nongnu.org; Fri, 07 Nov 2008 03:44:20 -0500 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1KyMx0-0004uw-BH for qemu-devel@nongnu.org; Fri, 07 Nov 2008 03:44:19 -0500 Received: from [199.232.76.173] (port=43220 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1KyMx0-0004up-1A for qemu-devel@nongnu.org; Fri, 07 Nov 2008 03:44:18 -0500 Received: from ecfrec.frec.bull.fr ([129.183.4.8]:37497) by monty-python.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1KyMx0-0008BW-1o for qemu-devel@nongnu.org; Fri, 07 Nov 2008 03:44:18 -0500 Received: from localhost (localhost [127.0.0.1]) by ecfrec.frec.bull.fr (Postfix) with ESMTP id DC5441A18E0 for ; Fri, 7 Nov 2008 09:44:13 +0100 (CET) Received: from ecfrec.frec.bull.fr ([127.0.0.1]) by localhost (ecfrec.frec.bull.fr [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 03730-06 for ; Fri, 7 Nov 2008 09:44:10 +0100 (CET) Received: from cyclope.frec.bull.fr (cyclope.frec.bull.fr [129.183.4.9]) by ecfrec.frec.bull.fr (Postfix) with ESMTP id A720D1A18E2 for ; Fri, 7 Nov 2008 09:44:10 +0100 (CET) Received: from [129.183.101.63] (frecb007144.frec.bull.fr [129.183.101.63]) by cyclope.frec.bull.fr (Postfix) with ESMTP id 6AC8127289 for ; Fri, 7 Nov 2008 09:44:07 +0100 (CET) Subject: Re: [Qemu-devel] [PATCH 0/4] qcow2: Improve cluster refcount update From: Laurent Vivier In-Reply-To: <49133D17.3050100@codemonkey.ws> References: <1225990556.6576.10.camel@frecb07144> <49133D17.3050100@codemonkey.ws> Content-Type: text/plain; charset=utf-8 Date: Fri, 07 Nov 2008 09:44:13 +0100 Message-Id: <1226047453.4046.4.camel@frecb07144> Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Reply-To: qemu-devel@nongnu.org List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Le jeudi 06 novembre 2008 =C3=A0 12:53 -0600, Anthony Liguori a =C3=A9cri= t : > Laurent Vivier wrote: > > This series of patches improves qcow2 performance with O_DIRECT > > when the qcow2 file is empty and we begin to fill it. > > =20 >=20 > I really dislike this series and any future series that does similar=20 > things. The real problem is that non-aligned accesses are so slow, but= =20 > there's no reason that they must be so slow. It's because we're doing=20 > synchronous IO operations instead of using posix-aio like we should. =20 > That's the real problem. Just doing a memory copy to an aligned buffer= =20 > is not going to cause that much performance delay. >=20 > We shouldn't be adding more cruft to the code base to avoid fixing the=20 > real problem. I think the current implementation of O_DIRECT needs to=20 > be rewritten. Like any good maintainer, you are becoming a tyrant... BTW, patch #1 is a clean-up (no change, only code move) and patch #2 doesn't align buffer, but try to minimise syscall (it improves performance with O_DSYNC too). Ignore patch #3 and #4 if you dislike them. Regards, Laurent --=20 ------------------ Laurent.Vivier@bull.net ------------------ "Tout ce qui est impossible reste =C3=A0 accomplir" Jules Verne "Things are only impossible until they're not" Jean-Luc Picard