Re: [PATCH 3/4] [BUGFIX] fix memcgroup LRU stat with THP

[Johannes Weiner <hannes@cmpxchg.org>]

On Fri, Jan 14, 2011 at 07:10:42PM +0900, KAMEZAWA Hiroyuki wrote:
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> 
> memroy cgroup's LRU stat should take care of size of pages because
> Transparent Hugepage inserts hugepage into LRU and zone counter
> is updeted based on the size of page.
> 
> If this value is the number wrong, memory reclaim will not work well.
> 
> Note: only head page of THP's huge page is linked into LRU.
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> ---
>  mm/memcontrol.c |   10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
> 
> Index: mmotm-0107/mm/memcontrol.c
> ===================================================================
> --- mmotm-0107.orig/mm/memcontrol.c
> +++ mmotm-0107/mm/memcontrol.c
> @@ -815,7 +815,10 @@ void mem_cgroup_del_lru_list(struct page
>  	 * removed from global LRU.
>  	 */
>  	mz = page_cgroup_zoneinfo(pc);
> -	MEM_CGROUP_ZSTAT(mz, lru) -= 1;
> +	if (!PageTransHuge(page))
> +		MEM_CGROUP_ZSTAT(mz, lru) -= 1;
> +	else
> +		MEM_CGROUP_ZSTAT(mz, lru) -= 1 << compound_order(page);

compound_order() returns 0 for !PG_head pages, that should do the
right thing without checking PageTransHuge(), right?