On Fri, 20 Sep 2013, Thomas Gleixner wrote: > Now when a cpu becomes isolated we stop the callback scheduling on > that cpu and assign it to the cpu with the smallest NUMA > distance. So that cpu will process the data for itself and for the > newly isolated cpu. That is not possible for many percpu threads since they rely on running on a specific cpu for optimization purposes. Running on a different processor makes these threads racy. What is needed is to be able to switch these things off and on. Something on a different cpu may monitor if processing on that specific cpu is needed or not but it cannot perform the vmstat updates. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>