Messages in this thread | | | Date | Thu, 21 Jan 2010 15:34:39 -0800 (PST) | From | David Rientjes <> | Subject | Re: [PATCH 5/7] Add /proc trigger for memory compaction |
| |
On Thu, 21 Jan 2010, Mel Gorman wrote:
> > It would be helpful to be able to determine what is "compactable" at the > > same time by adding both global and per-node "compact_order" tunables that > > would default to HUGETLB_PAGE_ORDER. > > Well, rather than having a separate tunable, writing a number to > /proc/sys/vm/compact could indicate the order if that trigger is now > working system-wide. Would that be suitable? >
Do you think you'll eventually find a need to call try_to_compact_pages() with a higher order than the one passed to the page allocator to limit "compaction thrashing" from fragmented frees to a zone where we're constantly compacting order-1 pages, for instance? I agree that memory compaction should always be used before direct reclaim for higher order allocations, but it may be more efficient to define a compact_min_order, tunable from userspace, that would avoid the need for constant order-1 compactions from subsequent page allocations.
If that's a possibility, we may find a need for "compact_order", now renamed "compact_min_order", outside of the explicit trigger.
| |