Messages in this thread | | | Date | Wed, 24 Nov 2004 08:40:20 -0200 | From | Marcelo Tosatti <> | Subject | Re: [PATCH]: 1/4 batch mark_page_accessed() |
| |
On Sun, Nov 21, 2004 at 01:12:50PM -0800, Andrew Morton wrote: > Nikita Danilov <nikita@clusterfs.com> wrote: > > > > Batch mark_page_accessed() (a la lru_cache_add() and lru_cache_add_active()): > > page to be marked accessed is placed into per-cpu pagevec > > (page_accessed_pvec). When pagevec is filled up, all pages are processed in a > > batch. > > > > This is supposed to decrease contention on zone->lru_lock. > > Looks sane, althought it does add more atomic ops (the extra > get_page/put_page). Some benchmarks would be nice to have.
I'll run STP benchmarks as soon as STP is working again. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |