From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:47231) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1SldAj-0006mm-EP for qemu-devel@nongnu.org; Mon, 02 Jul 2012 05:44:03 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1SldAe-0006xb-Dm for qemu-devel@nongnu.org; Mon, 02 Jul 2012 05:43:56 -0400 Received: from mx1.redhat.com ([209.132.183.28]:35769) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1SldAe-0006xM-5e for qemu-devel@nongnu.org; Mon, 02 Jul 2012 05:43:52 -0400 Message-ID: <4FF16D54.3020801@redhat.com> Date: Mon, 02 Jul 2012 12:43:48 +0300 From: Avi Kivity MIME-Version: 1.0 References: <20120702091813.GF8268@redhat.com> <4FF16A21.7050309@siemens.com> In-Reply-To: <4FF16A21.7050309@siemens.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] plan for device assignment upstream List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Jan Kiszka Cc: Marcelo Tosatti , Michael Tokarev , "qemu-devel@nongnu.org" , kvm , "Michael S. Tsirkin" On 07/02/2012 12:30 PM, Jan Kiszka wrote: > On 2012-07-02 11:18, Michael S. Tsirkin wrote: >> I've been thinking hard about Jan's patches for device >> assignment. Basically while I thought it makes sense >> to make all devices: assignment and not - behave the >> same and use same APIs for injecting irqs, Anthony thinks there is huge >> value in making irq propagation hierarchical and device assignment >> should be special cased. > > On the long term, we will need direct injection, ie. caching, to allow > making it lock-less. Stepping through all intermediate layers will cause > troubles, at least performance-wise, when having to take and drop a lock > at each stop. So we precalculate everything beforehand. Instead of each qemu_irq triggering a callback, calculating the next hop and firing the next qemu_irq, configure each qemu_irq array with a function that describes how to take the next hop. Whenever the configuration changes, recalculate all routes. For device assignment or vhost, we can have a qemu_irq_irqfd() which converts a qemu_irq to an eventfd. If the route calculations determine that it can be serviced via a real irqfd, they also configure it as an irqfd. Otherwise qemu configures a poll on this eventfd and calls the callback when needed. -- error compiling committee.c: too many arguments to function