| |||||||||||||
A big.LITTLE scheduler updateA big.LITTLE scheduler updatePosted Jun 12, 2012 22:33 UTC (Tue) by Tara_Li (subscriber, #26706)Parent article: A big.LITTLE scheduler update
QQ2 - Answer 3: We got into this mess by assuming symmetric multiprocessors would stay the rule for a long time. This is the tip of the iceberg, and we should be aiming for the most general solution we can come up with, as it's going to take a lot of refinement down the road to get right.
(Log in to post comments)
A big.LITTLE scheduler update Posted Jun 13, 2012 14:21 UTC (Wed) by PaulMcKenney (subscriber, #9624) [Link]
Indeed, there is a large amount of work ahead, and big.LITTLE is no doubt only the start. But I find it quite encouraging that small changes can produce meaningful improvements: This means that we can make use of asymmetric hardware now and improve its performance, energy efficiency, and ease of use over time.
But yes, there might well be very large changes required over the longer term. It should be fun! ;-)
|
|||||||||||||