From mboxrd@z Thu Jan  1 00:00:00 1970
From: Paul Brook <paul@codesourcery.com>
Subject: Re: [Qemu-devel] Re: [PATCH] virtio-blk: add SGI_IO passthru support
Date: Wed, 29 Apr 2009 12:37:20 +0100
Message-ID: <200904291237.21558.paul@codesourcery.com>
References: <20090427082606.GA32604@lst.de> <200904291211.20374.paul@codesourcery.com> <20090429112130.GA11241@lst.de>
Mime-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Cc: qemu-devel@nongnu.org, Anthony Liguori <anthony@codemonkey.ws>,
	kvm@vger.kernel.org, Rusty Russell <rusty@rustcorp.com.au>,
	Christian Borntraeger <borntraeger@de.ibm.com>,
	Hannes Reinecke <hare@suse.de>
To: Christoph Hellwig <hch@lst.de>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mail.codesourcery.com ([65.74.133.4]:57527 "EHLO
	mail.codesourcery.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1751099AbZD2LhX (ORCPT <rfc822;kvm@vger.kernel.org>);
	Wed, 29 Apr 2009 07:37:23 -0400
In-Reply-To: <20090429112130.GA11241@lst.de>
Content-Disposition: inline
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Wednesday 29 April 2009, Christoph Hellwig wrote:
> On Wed, Apr 29, 2009 at 12:11:19PM +0100, Paul Brook wrote:
> > Is this actually measurably faster, or just infinitesimally faster in
> > theory?
>
> On normal disks it's rather theoretical.  On high-end SSDs and arrays the
> impact is noticeable, mostly due to the additional latency.

How exactly does it introduce additional latency? A scsi command block is 
hardly large or complicated. Are you suggesting that a 16/32byte scsi command 
takes significantly longer to process than a 16byte virtio command 
descriptor? I'd expect any extra processing to be a small fraction of the 
host syscall latency, let alone the latency of the physical host adapter. It 
probably even fits on the same CPU cache line.

Paul