On Wed, Nov 06, 2019 at 12:51:30AM +0100, Paolo Bonzini wrote: > On 05/11/19 21:02, Peter Zijlstra wrote: > > On Tue, Nov 05, 2019 at 05:17:37PM +0100, Vitaly Kuznetsov wrote: > >> Virtualized guests may pick a different strategy to mitigate hardware > >> vulnerabilities when it comes to hyper-threading: disable SMT completely, > >> use core scheduling, or, for example, opt in for STIBP. Making the > >> decision, however, requires an extra bit of information which is currently > >> missing: does the topology the guest see match hardware or if it is 'fake' > >> and two vCPUs which look like different cores from guest's perspective can > >> actually be scheduled on the same physical core. Disabling SMT or doing > >> core scheduling only makes sense when the topology is trustworthy. > >> > >> Add two feature bits to KVM: KVM_FEATURE_TRUSTWORTHY_SMT with the meaning > >> that KVM_HINTS_TRUSTWORTHY_SMT bit answers the question if the exposed SMT > >> topology is actually trustworthy. It would, of course, be possible to get > >> away with a single bit (e.g. 'KVM_FEATURE_FAKE_SMT') and not lose backwards > >> compatibility but the current approach looks more straightforward. > > > > The only way virt topology can make any sense what so ever is if the > > vcpus are pinned to physical CPUs. > > This is a subset of the requirements for "trustworthy" SMT. You can have: > > - vCPUs pinned to two threads in the same core and exposed as multiple > cores in the guest Why the .... would one do anything like that? > - vCPUs from different guests pinned to two threads in the same core > > and that would be okay as far as KVM_HINTS_REALTIME is concerned, but > would still allow exploitation of side-channels, respectively within the > VM and between VMs. Hardly, RT really rather would not have SMT. SMT is pretty crap for determinism.
