From mboxrd@z Thu Jan  1 00:00:00 1970
From: Peter Zijlstra <peterz@infradead.org>
Subject: Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to
 support pv-ticketlock
Date: Wed, 19 Jan 2011 18:50:15 +0100
Message-ID: <1295459415.28776.146.camel@laptop>
References: <cover.1289940821.git.jeremy.fitzhardinge@citrix.com>
	<20110119164432.GA30669@linux.vnet.ibm.com>
	<20110119171239.GB726@linux.vnet.ibm.com>
	<20110119172306.GA2863@linux.vnet.ibm.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Cc: Mathieu, Xen-devel <xen-devel@lists.xensource.com>,
	Desnoyers <mathieu.desnoyers@polymtl.ca>,
	Nick Piggin <npiggin@kernel.dk>, Jeremy, Linux,
	Jeremy Fitzhardinge <jeremy@goop.org>, Beulich <JBeulich@novell.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>, Jan,
	Fitzhardinge <jeremy.fitzhardinge@citrix.com>, suzuki@in.ibm.com,
	Avi Kivity <avi@redhat.com>, kvm@vger.kernel.org,
	"H. Peter Anvin" <hpa@zytor.com>,
	=?ISO-8859-1?Q?Am=E9rico?= Wang <xiyou.wangcong@gmail.com>,
	Eric Dumazet <dada1@cosmosbay.com>,
	Virtualization <virtualization@lists.linux-foundation.org>
To: vatsa@linux.vnet.ibm.com
Return-path: <xen-devel-bounces@lists.xensource.com>
In-Reply-To: <20110119172306.GA2863@linux.vnet.ibm.com>
List-Unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xensource.com>
List-Help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-Subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
Sender: xen-devel-bounces@lists.xensource.com
Errors-To: xen-devel-bounces@lists.xensource.com
List-Id: kvm.vger.kernel.org

On Wed, 2011-01-19 at 22:53 +0530, Srivatsa Vaddagiri wrote:
> On Wed, Jan 19, 2011 at 10:42:39PM +0530, Srivatsa Vaddagiri wrote:
> > Add two hypercalls to KVM hypervisor to support pv-ticketlocks.
> >=20
> > KVM_HC_WAIT_FOR_KICK blocks the calling vcpu until another vcpu kicks i=
t or it
> > is woken up because of an event like interrupt.
>=20
> One possibility is to extend this hypercall to do a directed yield as wel=
l,=20
> which needs some more thought. Another issue that needs to be resolved wi=
th
> pv-ticketlocks is the impact on intra-VM fairness. A guest experiencing h=
eavy
> contention can keep yielding cpu, allowing other VMs to get more time tha=
n they
> deserve.

No, yield sucks, if you know the owner you can do actual PI.