|
|
|
|
|
|
|
|
|
|
xen-devel
[Xen-devel] Re: [PATCH] CPUIDLE: revise tsc-save/restore to avoid big ts
On 05/12/2008 11:30, "Wei, Gang" <gang.wei@xxxxxxxxx> wrote:
>> I tried extrapolating from t->stime_local_stamp, cpu_khz, and
>> t->local_tsc_stamp before I got into the current solution. It would still
>> bring accumulating skew, but in a slower increasing speed. I would like to
>> try it again with t->tsc_scale instead of cpu_khz. If it is works, it would
>> really be simpler. Allow me some time.
>
> Below patch should be what you expected. It will still bring continuously
> increasing tsc skew. If I pin all domains' vcpus on 1 pcpu, the skew is
> increasing faster.
I applied a modified version of this as c/s 18873. Primarily I removed
scale_time() and replaced with a simpler reciprocal function which is
applied to scale factors to invert them. Please take a look.
As I replied to Kevin, you can still have a scheme based on exploiting
constant_tsc if you like, but it needs to be enabled only when you really
have constant_tsc, and also subject to other comments I made in that email.
-- Keir
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
|
|
|
|