From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeremy Fitzhardinge Subject: Re: vsyscall support in 2.6.{32,33} Date: Tue, 21 Sep 2010 10:31:55 -0700 Message-ID: <4C98EC0B.3090005@goop.org> References: <4C98A9CD.1090900@redhat.com> <20100921130910.GQ2804@reaktio.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <20100921130910.GQ2804@reaktio.net> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: =?ISO-8859-1?Q?Pasi_K=E4rkk=E4inen?= Cc: Paolo Bonzini , Drew Jones , "xen-devel@lists.xensource.com" List-Id: xen-devel@lists.xenproject.org On 09/21/2010 06:09 AM, Pasi K=E4rkk=E4inen wrote: > On Tue, Sep 21, 2010 at 02:49:17PM +0200, Paolo Bonzini wrote: >> Besides the obvious fact that the required hypercall is not supported = by =20 >> any released hypervisor, is there any other reason why the xen vsyscal= l =20 >> code is not in the pvops/stable-2.6.{32,33}.x branches? >> >> Thanks! >> > I guess Jeremy didn't remember to merge it yet? :) > Or is there some problem with it? > The main problem is that I'm not sure that the Xen clock algorithm can give strict enough monotonicity for usermode use (ie, can't meet the expected behaviour of clock_gettime()), and its not at all clear how to fix it. I could prevent non-monotonicity within one process, but not between processes. J