From mboxrd@z Thu Jan 1 00:00:00 1970 From: Rusty Russell Subject: Re: [PATCH 2/3] virtio: indirect ring entries (VIRTIO_RING_F_INDIRECT_DESC) Date: Tue, 12 May 2009 23:53:53 +0930 Message-ID: <200905122353.54424.rusty@rustcorp.com.au> References: <1229620222-22216-1-git-send-email-markmc@redhat.com> <200905041149.00724.rusty@rustcorp.com.au> <1242061838.25337.8.camel@blaa> Mime-Version: 1.0 Content-Type: Text/Plain; charset="utf-8" Content-Transfer-Encoding: 7bit Cc: dlaor@redhat.com, netdev@vger.kernel.org, Dor Laor , Avi Kivity , virtualization@lists.linux-foundation.org To: Mark McLoughlin Return-path: Received: from ozlabs.org ([203.10.76.45]:52526 "EHLO ozlabs.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751710AbZELOYH (ORCPT ); Tue, 12 May 2009 10:24:07 -0400 In-Reply-To: <1242061838.25337.8.camel@blaa> Content-Disposition: inline Sender: netdev-owner@vger.kernel.org List-ID: On Tue, 12 May 2009 02:40:38 am Mark McLoughlin wrote: > > Blocked from doing the simpler thing, we've decided to go with a layer > > of indirection. But the patch is simple and clean, so there's nothing > > fundamental to object to. > > Still have one FIXME in the patch worth looking at - at what point > should we use an indirect entry rather than consuming N entries? OK, I've applied these as is. I'm doing some virtio net benchmarking (under lguest); I'll see if I can get a reasonable figure. I don't think there's an obvious right answer; it depends how many more packets are coming as well as how many descriptors each will use. Thanks, Rusty.