[Lguest] [BUG + PATCH/Bugfix] x86/lguest: fix pgdir pmd index calculation

Rusty Russell rusty at rustcorp.com.au
Tue Mar 4 23:06:57 EST 2008


On Saturday 01 March 2008 06:58:38 Ingo Molnar wrote:
> * Ahmed S. Darwish <darwish.07 at gmail.com> wrote:
> > On the lguest _guest_ side, 'cpu_has_tsc' is _always_ false (due to
> > lguest using his own clocksource ?), thus a guest with a pentium+ cpu
> > always panics with:
>
> does the "RDTSC" instruction work in an lguest guest? If not, then the
> lguest kernel is correct in not exposing it - and then the solution is
> to build a non-TSC guest kernel. Does the patch below help?
>
> but if the RDTSC instruction does work in an lguest guest, then the
> proper approach would be to expose it in the CPU features.

Yes, and after some investigation, I came up with two patches.  I'll queue 
both to you if that's ok, since they're dependent.

When lguest was being developed we didn't have a get_tsc_khz hook, so I had to 
kill the TSC cpuid bit.  Nowdays, the code can actually be much more natural: 
the Host tells us the cpu khz, and if it's zero, we fall back to the 
dumb "lguest" clock.

Thanks,
Rusty.



More information about the Lguest mailing list