Re: [PATCH 0/2] Expose VMFUNC to the nested hypervisor

2017-06-30 Thread Paolo Bonzini
- Original Message - > From: "Jim Mattson" > To: "Bandan Das" > Cc: "kvm list" , "Paolo Bonzini" , > "LKML" > Sent: Friday, June 30, 2017 7:06:43 PM > Subject: Re: [PATCH 0/2] Expose VMFUNC to the nested hypervisor

Re: [PATCH 0/2] Expose VMFUNC to the nested hypervisor

2017-06-30 Thread Bandan Das
Jim Mattson writes: > Isn't McAfee DeepSAFE defunct? Are there any other consumers of EPTP > switching? I don't know of any real users but I think we should be providing this functionality to the L1 hypervisor :) IIRC, Xen lets you use EPTP switching as part of VM introspection ? Bandan >

Re: [PATCH 0/2] Expose VMFUNC to the nested hypervisor

2017-06-30 Thread Jim Mattson
Isn't McAfee DeepSAFE defunct? Are there any other consumers of EPTP switching? On Thu, Jun 29, 2017 at 4:29 PM, Bandan Das wrote: > These patches expose eptp switching/vmfunc to the nested hypervisor. Testing > with > kvm-unit-tests seems to work ok. > > If the guest hypervisor enables vmfunc/e

[PATCH 0/2] Expose VMFUNC to the nested hypervisor

2017-06-29 Thread Bandan Das
These patches expose eptp switching/vmfunc to the nested hypervisor. Testing with kvm-unit-tests seems to work ok. If the guest hypervisor enables vmfunc/eptp switching, a "shadow" eptp list address page is written to the VMCS. Initially, it would be unpopulated which would result in a vmexit wit