From mboxrd@z Thu Jan 1 00:00:00 1970 From: George Dunlap Subject: Re: [PATCH for-4.5 v6 05/16] tools: Add vmware_port support Date: Wed, 24 Sep 2014 17:24:27 +0100 Message-ID: <5422F03B.30900@eu.citrix.com> References: <1411236447-7435-1-git-send-email-dslutz@verizon.com> <1411236447-7435-6-git-send-email-dslutz@verizon.com> <1411393315.18331.104.camel@kazak.uk.xensource.com> <54204FA5.7020104@citrix.com> <5420930A.2020300@terremark.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <5420930A.2020300@terremark.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: Don Slutz , Andrew Cooper , Ian Campbell Cc: Tim Deegan , Kevin Tian , Keir Fraser , Jun Nakajima , Stefano Stabellini , Ian Jackson , Eddie Dong , xen-devel@lists.xen.org, Aravind Gopalakrishnan , Jan Beulich , Boris Ostrovsky , Suravee Suthikulpanit List-Id: xen-devel@lists.xenproject.org On 09/22/2014 10:22 PM, Don Slutz wrote: > On 09/22/14 12:34, Andrew Cooper wrote: >> On 22/09/14 14:41, Ian Campbell wrote: >>> On Sat, 2014-09-20 at 14:07 -0400, Don Slutz wrote: >>>> This new libxl_domain_create_info field is used to set >>>> XEN_DOMCTL_CDF_vmware_port for the xc_domain_create() routine. >>> Does this really need to be a CDF, rather than a domctl/hvm param? >> I have made the argument that many things which are currently HVM Params >> should be CDF, as they absolutely should be set and immutable for the >> entire lifetime of the domain. >> >> From recollection, we have had several XSAs in the past which are >> directly attributable to the toolstack or guest being able to play with >> an (insufficiently locked down) HVM param after boot. >> >> Using a CDF avoids potential issues along these lines. > > It also allow setting up v->arch.hvm_vmx.exception_bitmap at > the right time. domctl/hvm params are setup much latter in > the life of a domain. Isn't that already modified on a cr change (a la vmx_update_guest_cr())? Or did you mean the SVM side? I'm not making an argument either way (although at the moment I'm more sympathetic to Andy's view), just questioning whether setting the exit flag is that much of an argument one way or another. -George