From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1760713Ab3BNPcL (ORCPT ); Thu, 14 Feb 2013 10:32:11 -0500 Received: from mail-oa0-f51.google.com ([209.85.219.51]:56894 "EHLO mail-oa0-f51.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758022Ab3BNPcJ (ORCPT ); Thu, 14 Feb 2013 10:32:09 -0500 Message-ID: <511D0374.9070202@codemonkey.ws> Date: Thu, 14 Feb 2013 09:32:04 -0600 From: Anthony Liguori User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:11.0) Gecko/20120329 Thunderbird/11.0.1 MIME-Version: 1.0 To: Pekka Enberg CC: Paolo Bonzini , Ingo Molnar , Linus Torvalds , "H. Peter Anvin" , Linux Kernel Mailing List , Randy Dunlap , Thomas Gleixner , David Rientjes , David Woodhouse , Greg Kroah-Hartman , Sasha Levin , "H. Peter Anvin" , Michal Marek , Stephen Rothwell Subject: Re: kvmtool tree (Was: Re: [patch] config: fix make kvmconfig) References: <20130211122654.GA5802@gmail.com> <20130211172807.GA9716@gmail.com> <20130211175856.GC9716@gmail.com> <20130212095249.GC20039@gmail.com> <511B4D76.50504@redhat.com> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 02/13/2013 02:56 AM, Pekka Enberg wrote: > On Wed, Feb 13, 2013 at 10:23 AM, Paolo Bonzini wrote: >> Il 12/02/2013 10:52, Ingo Molnar ha scritto: >>> Check the list I gave (unmodified): >>> >>> "- Pekka listed new virtio drivers that were done via tools/kvm. >> >> vhost-scsi got in first in tools/kvm, but out-of-tree patches had >> existed for QEMU for more than a year. It was developed with QEMU. > > I think Ingo confused virtio and vhost. IIRC, Asias developed > vhost-blk using tools/kvm. We've done extensive performance analysis of vhost-blk and it's not any faster than a userspace solution. This is why it's still not in mainline. This wasn't noticed with tools/kvm because it's block layer is too simplistic. In order to get to the point where we're able to do this well in userspace in QEMU took tons of bug fixes to the kernel and added features (like pread64/pwrite64). That all happened without QEMU being in the kernel. Regards, Anthony Liguori