Subj : Re: Hazard Pointers w/o memory barrier.
To   : comp.programming.threads
From : Chris Thomasson
Date : Wed Apr 20 2005 11:45 am

>>> Pretty quick.  That's what I spent part of yesterday doing.  Verifying
>>> that
>>> Windows had a virtual thread timer that was updated on the fly.
>>
>> http://blog.kalmbachnet.de/?postid=28
>>
>
> I was wondering about that.  It only happens with Sleep().  If you do
> SwitchToThread() the times are updated.

Humm, need to try that.

You could probably also detect quiescent states by counting the 
context-switches per-thread using the performance api's...

.