On 02/24/2010 05:52 PM, Andrea Arcangeli wrote:
As sysctl that control defrag there is only one right now and it turns defrag on and off. We could make it more finegrined and have two files, one for the page faults in transparent_hugepage/defrag as always|madvise|never, and one yes|no in transparent_hugepage/khugepaged/defrag but that may be overdesign... I'm not sure really when and how to invoke memory compaction, so having that maximum amount of knobs really is only requires if we can't came up with an optimal design.
The knobs may be useful for testing what the best default behaviour would be, as well as for workloads with special needs (eg. HPC vs. workloads with many short lived small processes). On systems with many CPU cores, it would also be fairly easy to do some background memory compaction on an idle CPU core, so we have huge pages available for when processes need them... There are a lot of unknowns here still, so we probably need to stay flexible for a while to come. The only thing we do know is that the hugepages can really speed up some workloads. -- 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/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>