Re: RFC: VMX: initialize TSC offset relative to vm creation time

[Avi Kivity <avi@redhat.com>]

David S. Ahern wrote:
> Has anything changed "recently" with the TSC code? Recently here being
> the past 2 months since you first crafted the patch. I ask because in
> the past few runs based on kvm.git trees (e.g., as recently as a pull on
> 10/26), this tsc offset patch no longer fixes the problem.
>
> The following one does fix the problem with kvm.git pulled on 10/26/08:
>
> diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
> index 64e2439..d5da717 100644
> --- a/arch/x86/kvm/vmx.c
> +++ b/arch/x86/kvm/vmx.c
> @@ -860,7 +860,7 @@ static void guest_write_tsc(u64 guest_tsc)
>         u64 host_tsc;
>
>         rdtscll(host_tsc);
> -       vmcs_write64(TSC_OFFSET, guest_tsc - host_tsc);
> +       vmcs_write64(TSC_OFFSET, 0);
>  }
>   

That's a bit heavy handed, it doesn't start he guest tsc from zero and 
doesn't allow the guest to adjust tsc.

But it does work for the case the tscs are synced.

-- 
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.