From mboxrd@z Thu Jan 1 00:00:00 1970 From: Konrad Rzeszutek Wilk Subject: Re: [PATCH RFC] xenbus_client: extend interface to suppurt multi-page ring Date: Tue, 20 Aug 2013 16:07:31 -0400 Message-ID: <20130820200731.GB12737@phenom.dumpdata.com> References: <1369922760-11800-1-git-send-email-wei.liu2@citrix.com> <20130530140912.GI4394@zion.uk.xensource.com> <20130820171158.GE11652@phenom.dumpdata.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Wei Liu Cc: boris.ostrovsky@oracle.com, Wei Liu , David Vrabel , "xen-devel@lists.xen.org" List-Id: xen-devel@lists.xenproject.org On Tue, Aug 20, 2013 at 06:41:29PM +0100, Wei Liu wrote: > On Tue, Aug 20, 2013 at 6:11 PM, Konrad Rzeszutek Wilk > wrote: > > On Thu, May 30, 2013 at 03:09:12PM +0100, Wei Liu wrote: > >> On Thu, May 30, 2013 at 03:06:00PM +0100, Wei Liu wrote: > >> > Originally Xen PV drivers only use single-page ring to pass along > >> > information. This might limit the throughput between frontend and > >> > backend as they can easily run out of ring slots if hardware is fast. > >> > > >> > The patch extends Xenbus driver to support multi-page ring, which in > >> > general should improve throughput. Changes to various frontend / backend > >> > to adapt to the new interface are also included. > >> > > >> > Affected Xen drivers: > >> > * blkfront/back > >> > * netfront/back > >> > * pcifront/back > >> > > >> > The interface is documented, as before, in xenbus_client.c. > >> > > >> > >> Please note that this patch won't apply cleanly on upstream kernel as it > >> relies on some patches that are not in tree. > > > > OK. > >> > >> I mainly want to collect views on the interface. Also in the HVM case I > >> use vmap, I would like to know if I get it right. > > > > This patches looks familiar that I had seen some time ago (maybe I even > > backported it for PV?). Anyhow the PV part looks OK. > > > > The HVM - I don't see why we cannot do vmap. The nature of the ring is that > > it stays mapped for a very long time - so using vmap is OK. That is, the > > issues with vmap is that it takes a bit to setup and teardown - and that > > is OK with us. > > > > If you would like to rebase it on top of v3.11-rc5 and repost it I don't > > see why we can't have it for v3.12. David, Boris? > > > > I think the above implementation has a problem in the HVM case, but I > cannot recall what. :-( Heh. Well does it work? > > If you're generally happy with the idea, I can polish this patch and > target it for 3.12. Yes sure. We still have some time before I need to clamp down. > > Wei. > > >> > >> > >> Thanks > >> Wei. > >> > >> _______________________________________________ > >> Xen-devel mailing list > >> Xen-devel@lists.xen.org > >> http://lists.xen.org/xen-devel > >> > > > > _______________________________________________ > > Xen-devel mailing list > > Xen-devel@lists.xen.org > > http://lists.xen.org/xen-devel