From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Michael S. Tsirkin" Subject: Re: [PATCH 2/5] virtio: support unlocked queue kick Date: Tue, 1 Nov 2011 16:40:45 +0200 Message-ID: <20111101144045.GA15433@redhat.com> References: <20111005195403.407628164@bombadil.infradead.org> <20111005195529.964397366@bombadil.infradead.org> <87r52qgaf3.fsf@rustcorp.com.au> <20111006131828.GC19023@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Christoph Hellwig , Chris Wright , Jens Axboe , Stefan Hajnoczi , kvm@vger.kernel.org, linux-kernel@vger.kernel.org To: Rusty Russell Return-path: Content-Disposition: inline In-Reply-To: <20111006131828.GC19023@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org On Thu, Oct 06, 2011 at 11:18:28AM -0200, Michael S. Tsirkin wrote: > On Thu, Oct 06, 2011 at 12:15:36PM +1030, Rusty Russell wrote: > > On Wed, 05 Oct 2011 15:54:05 -0400, Christoph Hellwig wrote: > > > Split virtqueue_kick to be able to do the actual notification outside the > > > lock protecting the virtqueue. This patch was originally done by > > > Stefan Hajnoczi, but I can't find the original one anymore and had to > > > recreated it from memory. Pointers to the original or corrections for > > > the commit message are welcome. > > > > An alternative to this is to update the ring on every virtqueue_add_buf. > > MST discussed this for virtio_net, and I think it's better in general. > > > > The only reason that it wasn't written that way originally is that the > > barriers make add_buf slightly more expensive. > > > > Cheers, > > Rusty. > > With event index, I'm not sure that's enough to make the kick lockless > anymore. Hmm, any comment on this? These have been benchmarked to give a sizeable speedup, so I'm thinking it's better to take the patches as is, if someone has the inclination to redo the work with an atomic virtqueue_add_buf, that can be applied on top. > > -- > MST