From mboxrd@z Thu Jan  1 00:00:00 1970
From: Christoph Hellwig <hch@lst.de>
Subject: Re: [PATCH, RFC] virtio_blk: add cache flush command
Date: Tue, 12 May 2009 09:26:27 +0200
Message-ID: <20090512072627.GC5627@lst.de>
References: <20090511083908.GB20082@lst.de> <4A083B7C.1000703@codemonkey.ws> <20090511154046.GA4226@lst.de> <4A085471.40809@codemonkey.ws>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Christoph Hellwig <hch@lst.de>,
	Rusty Russell <rusty@rustcorp.com.au>, kvm@vger.kernel.org
To: Anthony Liguori <anthony@codemonkey.ws>
Return-path: <kvm-owner@vger.kernel.org>
Received: from verein.lst.de ([213.95.11.210]:44928 "EHLO verein.lst.de"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1752721AbZELH0g (ORCPT <rfc822;kvm@vger.kernel.org>);
	Tue, 12 May 2009 03:26:36 -0400
Content-Disposition: inline
In-Reply-To: <4A085471.40809@codemonkey.ws>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Mon, May 11, 2009 at 11:38:09AM -0500, Anthony Liguori wrote:
> >Right now it doesn't, but it probably should.
> >  
> 
> So then with cache=writeback, fsync behaves itself but O_DIRECT writes 
> do not.

Right now O_DIRECT does not do an explicit cache flush, but due to the
way barriers are implemented in Linux we do get the cache flush as part
of the metadata updates after I/O completion.

> fsync is pretty crappy on ext3 default configs.  I'm concerned that this 
> could be considered a DoS by a malicious guest.  If it sat in a T_FLUSH 
> loop, it would potentially bring your system to a crawl, no?

It's exactly the same effect as a regular user doing fsync in a loop,
whatever that causes on ext3.