| |||||||||||||
High- (but not too high-) resolution timeoutsHigh- (but not too high-) resolution timeoutsPosted Sep 6, 2008 15:23 UTC (Sat) by kbob (guest, #1770)Parent article: High- (but not too high-) resolution timeouts
"This value can be adjusted via the prctl() system call. The default value is set to 50 microseconds - approximate to a certain degree, but still far more accurate than the timeouts in current kernels"
Why not set the default value to one jiffy? That would maintain compatibility with older kernels, and the longer default value would result in fewer wakeups for programs that haven't called prctl().
(Log in to post comments)
High- (but not too high-) resolution timeouts Posted Sep 6, 2008 17:39 UTC (Sat) by arjan (subscriber, #36785) [Link]
one of the issues is that the old code had a *maximum* of a 1 millisecond, but the average was more like 500 usec.
The other thing is... with the code in the patchkit, the behavior in terms of power isn't all that bad (after all, userland isn't poll()ing like crazy anymore; all that got fixed with powertop).
|
|||||||||||||