On Wed, Mar 02, 2011 at 07:33:13AM +0900, Minchan Kim wrote: > Sorry for bothering you but I think you get the data. > It helps someone in future very much to know why we determined to > remove the feature at that time and they should do what kinds of > experiment to prove it has a benefit to add compaction in kswapd > again. This is a benchmark I'm unsure if it's ok to publish results but it should be possible to simulate it with a device driver. Arthur provided kswapd load usage data too, so I hope that's enough. My other patch (compaction-kswapd-3) is way better than current logic and retains compaction in kswapd. That shows slightly higher kswapd utilization with Arthur's multimedia workload, and a bit worse performance on the network benchmark. So I thought it was better to go with the fastest potion as long as we don't have a logic that uses compaction and shows improved performance and lower latency than with no compaction at all in kswapd. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>