Re: time interpolation hooks

David Mosberger (davidm@napali.hpl.hp.com)
Mon, 19 May 2003 11:35:33 -0700


>>>>> On Mon, 19 May 2003 17:42:03 +0000, Arjan van de Ven <arjanv@redhat.com> said:

Arjan> On Mon, May 19, 2003 at 10:34:43AM -0700, David Mosberger wrote:

>> struct time_interpolator {
>> void (*update_wall_time) (long delta_nsec);
>> void (*reset_wall_time) (long delta_nsec);
>> unsigned long frequency; /* frequency in counts/second */
>> unsigned long drift; /* drift in parts-per-million (?) */
>> };

Arjan> probably also a "get interpolated value" kind of thing ?

Yes, of course. Not a good idea to write an email in a hurry.

Arjan> other than that it seems to match what I had in mind.
Arjan> For the score we may need something creative; I'm not sure all timers
Arjan> have a defined drift, otoh parts-per-million seems to be the
Arjan> standard mechanism of reporting this.

What could you possibly do about a timer with an unknown drift?
Perhaps give preference to higher frequency, even if the drift is
unknown? In any case, I suppose we can fine-tune the policy
afterwards.

Andrew, I assume it's OK with you if I update the ia64 code to the
proposed interface and then send you an updated patch?

--david
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/