From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jeremy Fitzhardinge <jeremy@goop.org>
Subject: Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
Date: Mon, 24 Jan 2011 09:49:58 -0800
Message-ID: <4D3DBBC6.1000407@goop.org>
References: <cover.1289940821.git.jeremy.fitzhardinge@citrix.com> <20110119164432.GA30669@linux.vnet.ibm.com> <20110119171239.GB726@linux.vnet.ibm.com> <1295457672.28776.144.camel@laptop> <4D373340.60608@goop.org> <20110120115958.GB11177@linux.vnet.ibm.com> <4D38774B.6070704@goop.org> <20110121140208.GA13609@linux.vnet.ibm.com> <4D399CBD.10506@redhat.com> <20110122061417.GA7258@linux.vnet.ibm.com> <4D3AEF7A.1050709@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Cc: vatsa@linux.vnet.ibm.com, Peter Zijlstra <peterz@infradead.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Nick Piggin <npiggin@kernel.dk>,
	Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>,
	=?UTF-8?B?QW3DqXJpY28gV2FuZw==?= <xiyou.wangcong@gmail.com>,
	Eric Dumazet <dada1@cosmosbay.com>,
	Jan Beulich <JBeulich@novell.com>, Avi Kivity <avi@redhat.com>,
	Xen-devel <xen-devel@lists.xensource.com>,
	"H. Peter Anvin" <hpa@zytor.com>,
	Linux Virtualization <virtualization@lists.linux-foundation.org>,
	Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>,
	kvm@vger.kernel.org, suzuki@in.ibm.com
To: Rik van Riel <riel@redhat.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from claw.goop.org ([74.207.240.146]:42383 "EHLO claw.goop.org"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1751813Ab1AXRuG (ORCPT <rfc822;kvm@vger.kernel.org>);
	Mon, 24 Jan 2011 12:50:06 -0500
In-Reply-To: <4D3AEF7A.1050709@redhat.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On 01/22/2011 06:53 AM, Rik van Riel wrote:
> The main question that remains is whether the PV ticketlocks are
> a large enough improvement to also merge those.  I expect they
> will be, and we'll see so in the benchmark numbers.

The pathological worst-case of ticket locks in a virtual environment
isn't very hard to hit, so I expect they'll make a huge difference
there.   On lightly loaded systems (little or no CPU overcommit) then
they should be close to a no-op.

    J