From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH 01/13] iommu bitmap insteads of iommu pointer in dmar_domain Date: Fri, 05 Dec 2008 18:42:05 +0200 Message-ID: <493959DD.2000109@redhat.com> References: <715D42877B251141A38726ABF5CABF2C018BF05987@pdsmsx503.ccr.corp.intel.com> <1228410746.3732.135.camel@blaa> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: "Han, Weidong" , "Woodhouse, David" , "'Jesse Barnes'" , "Yu, Fenghua" , "'iommu@lists.linux-foundation.org'" , "'kvm@vger.kernel.org'" To: Mark McLoughlin Return-path: Received: from mx2.redhat.com ([66.187.237.31]:53017 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751834AbYLEQmk (ORCPT ); Fri, 5 Dec 2008 11:42:40 -0500 In-Reply-To: <1228410746.3732.135.camel@blaa> Sender: kvm-owner@vger.kernel.org List-ID: Mark McLoughlin wrote: > Also, I wasn't sure at first what you meant by "native VT-d" ... you > mean DMA-API VT-d usage as opposed to KVM device assignment usage, > right? Perhaps we need a better term for that distinction. > > dma-api is request-oriented (the API is called to set up and tear down a single dma request), while the kvm usage is address space oriented (the API is called to set up, maintain, and tear down an address space, over which many dma requests can run). Perhaps someone can come up with names that reflect that. -- I have a truly marvellous patch that fixes the bug which this signature is too narrow to contain.