Re: [PATCH 2/2] memcg: add mlock statistic in memory.stat

[KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>]

On Fri, Apr 27, 2012 at 3:14 PM, Ying Han <yinghan@google.com> wrote:
> We have the nr_mlock stat both in meminfo as well as vmstat system wide, this
> patch adds the mlock field into per-memcg memory stat. The stat itself enhances
> the metrics exported by memcg since the unevictable lru includes more than
> mlock()'d page like SHM_LOCK'd.
>
> Why we need to count mlock'd pages while they are unevictable and we can not
> do much on them anyway?
>
> This is true. The mlock stat I am proposing is more helpful for system admin
> and kernel developer to understand the system workload. The same information
> should be helpful to add into OOM log as well. Many times in the past that we
> need to read the mlock stat from the per-container meminfo for different
> reason. Afterall, we do have the ability to read the mlock from meminfo and
> this patch fills the info in memcg.
>
> Note:
> Here are the places where I didn't add the hook:
> 1. in the mlock_migrate_page() since the owner of oldpage and newpage is the same.
> 2. in the freeing path since page shouldn't get to there at the first place.

Looks good. (even though I don't like unreliable mlock statistics)
 Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  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@kvack.org"> email@kvack.org </a>