Yes, that does indeed fix it. Thanks very much!
Now, onward to scaling! (Necessary when the PV kernel
boots on "tsc safe" hardware but then migrates.)
Thanks again!
Dan
> -----Original Message-----
> From: Keir Fraser [mailto:keir.fraser@xxxxxxxxxxxxx]
> Sent: Saturday, November 21, 2009 2:45 PM
> To: Dan Magenheimer; Jeremy Fitzhardinge
> Cc: Xen-Devel (E-mail)
> Subject: Re: Bizarre pv kernel ultra-high frequency rdtsc?!?
>
>
> On 20/11/2009 23:45, "Dan Magenheimer"
> <dan.magenheimer@xxxxxxxxxx> wrote:
>
> > - _u.tsc_timestamp = t->stime_local_stamp;
> > - _u.system_time = t->stime_local_stamp;
> > + _u.tsc_timestamp = t->stime_local_stamp - BIG_OFFSET;
> > + _u.system_time = t->stime_local_stamp - BIG_OFFSET;
>
> Ah, the line above is the bug. You should be subtracting
> BIG_OFFSET from
> _u.tsc_timestamp only. Leave _u.system_time as it is. That'll
> fix all your
> observed problems.
>
> -- Keir
>
>
>
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|