From mboxrd@z Thu Jan  1 00:00:00 1970
From: Avi Kivity <avi@qumranet.com>
Subject: Re: [PATCH] KVM: PCIPT: VT-d: fix guest unmap
Date: Fri, 20 Jun 2008 21:59:48 +0300
Message-ID: <485BFE24.5080100@qumranet.com>
References: <1213729526-10410-1-git-send-email-benami@il.ibm.com> <1213729526-10410-2-git-send-email-benami@il.ibm.com> <1213729526-10410-3-git-send-email-benami@il.ibm.com> <48582CD0.5060109@codemonkey.ws>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: benami@il.ibm.com, amit.shah@qumranet.com, weidong.han@intel.com,
	muli@il.ibm.com, raharper@us.ibm.com, kvm@vger.kernel.org
To: Anthony Liguori <anthony@codemonkey.ws>
Return-path: <kvm-owner@vger.kernel.org>
Received: from il.qumranet.com ([212.179.150.194]:24274 "EHLO il.qumranet.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1753002AbYFTS7t (ORCPT <rfc822;kvm@vger.kernel.org>);
	Fri, 20 Jun 2008 14:59:49 -0400
In-Reply-To: <48582CD0.5060109@codemonkey.ws>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

Anthony Liguori wrote:
> I think the current VT-d code needs some reworking.
>
> We should build the table as the shadow page table gets built.  We 
> should suppress iotlb flushes unless the table is actually being updated.
>

We can't, since we need the iommu tables populated before we issue any dma.

Perhaps we want something like MAP_POPULATE for shadow, which would then 
affect the iommu tables.  Userspace would then do:

mlock()
... set up device assignment
ioctl(..., KVM_POPULATE)


-- 
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.