From mboxrd@z Thu Jan 1 00:00:00 1970 From: DuanZhenzhong Subject: Re: [PATCH 3/3 V3] XSA-60 security hole: cr0.cd handling Date: Wed, 23 Oct 2013 16:48:14 +0800 Message-ID: <52678D4E.1000603@oracle.com> References: <5266AE0D02000078000FCB84@nat28.tlf.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <5266AE0D02000078000FCB84@nat28.tlf.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Jan Beulich Cc: Jinsong Liu , Tim Deegan , Keir Fraser , "suravee.suthikulpanit@amd.com" , Andrew Cooper , Eddie Dong , "xen-devel@lists.xen.org" , Will Auld , Jun Nakajima , "sherry.hurwitz@amd.com" List-Id: xen-devel@lists.xenproject.org Jan Beulich wrote: >>>> On 21.10.13 at 17:55, "Liu, Jinsong" wrote: >>>> >> From 4ff1e2955f67954e60562b29a00adea89e5b93ae Mon Sep 17 00:00:00 2001 >> From: Liu Jinsong >> Date: Thu, 17 Oct 2013 05:49:23 +0800 >> Subject: [PATCH 3/3 V3] XSA-60 security hole: cr0.cd handling >> >> This patch solves XSA-60 security hole: >> 1. For guest w/o VT-d, and for guest with VT-d but snooped, Xen need >> do nothing, since hardware snoop mechanism has ensured cache coherency. >> >> 2. For guest with VT-d but non-snooped, cache coherency can not be >> guaranteed by h/w snoop, therefore it need emulate UC type to guest: >> 2.1). if it works w/ Intel EPT, set guest IA32_PAT fields as UC so that >> guest memory type are all UC. >> 2.2). if it works w/ shadow, drop all shadows so that any new ones would >> be created on demand w/ UC. >> >> This patch also fix a bug of shadow cr0.cd setting. Current shadow has a >> small window between cache flush and TLB invalidation, resulting in possilbe >> cache pollution. This patch pause vcpus so that no vcpus context involved >> into the window. >> >> Signed-off-by: Liu Jinsong >> > > This looks fine to me now, but will need acks/reviews at least from > - Keir (whose blessing of the pausing construct I'd like to have even > if this didn't involve changing non-x86 files) > - one of the VMX maintainers > - one or both of Tim and Andrew > > And of course I'd really appreciate if Oracle could arrange for > testing this, to confirm their performance problem is also gone with > this. > I am try finding an env to test it. I'll reply after test. zduan