On Fri, 2022-02-25 at 16:21 +0000, Sean Christopherson wrote: > > I also have questions about the use case for the above patch.... if > > this is a clean boot and you're just starting to host guests, surely we > > can wait for the time it takes for the TSC synchronization to complete? > > KVM is built into the kernel in their case, the vmx_init() => kvm_init() gets > automatically called during boot. The VMs aren't started until well after > synchronization has completed, but KVM has already snapshotted the "bad" value. Gotcha. So even when we put my patch in front, to snapshot a value into kvm->arch.default_tsc_khz, that's happening later at VM creation time so should also be snapshotting the *good* value. And at least if it snapshots the bad value, all the vCPUs will be *consistent*.
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
