On Mon, May 23, 2016 at 06:55:06PM -0300, Marcelo Tosatti wrote:
> On Thu, Apr 21, 2016 at 08:21:25PM +0300, Roman Kagan wrote:
> > From: Paolo Bonzini <pbonzini@xxxxxxxxxx>
> >
> > The test checks the relative precision of the reference TSC page
> > and the time reference counter.
> >
> > Signed-off-by: Paolo Bonzini <pbonzini@xxxxxxxxxx>
> > [Adjust types to pass printf type checks - Roman Kagan]
> > Signed-off-by: Roman Kagan <rkagan@xxxxxxxxxxxxx>
> [...]
> > + if (t >= end) {
> > + long long ref = rdmsr(HV_X64_MSR_TIME_REF_COUNT);
> > + if (i == 0)
> > + printf("Time reference MSR drift: %lld\n\n", ref - end);
> > + ok[i] &= (ref - end) > -5 && (ref - end) < 5;
>
> This is prone to fail: guest can be scheduled at "***" above and
> test will fail.
>
> The [-5,5] comes from the standard? What the standard dictates?
No standard of course, it's arbitrary.
I tend to think a better test would be to alternate clock reads from the
MSR and the TSC page and check that neither goes ahead of the other.
Roman.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
