From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Jan Beulich" Subject: RE: granting access to MSI-X table and pending bit array Date: Thu, 05 Aug 2010 18:44:28 +0100 Message-ID: <4C5B068D0200007800061BF0@vpn.id2.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Return-path: Content-Disposition: inline List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: yunhong.jiang@intel.com, konrad.wilk@oracle.com Cc: xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org >>> "Jiang, Yunhong" 07/15/10 10:24 AM >>> >Will the small window for qemu exists still? QEMU may setup the mapping = before the msix_capability_init(). Yes, I think so, as Dom0 still is permitted to map this area writable = (revoking this requires a qemu fix to be committed at the same time, or a = prerequisite fix to make qemu map this area read-only, which I think = doesn't fit well within the MMIO model used there). >BTW, I'm not sure if mapping as RO this will solve the sharing issue = (i.e. the PBA or table entry share with other resource). Not if the device meets the standard's requirement of these two tables = being allowed to share a page between them, but not with any other = resource. Jan