mm: memcontrol: drain memcg stock on force_empty
diff mbox

Message ID 20180507201651.165879-1-shakeelb@google.com
State New
Headers show

Commit Message

Shakeel Butt May 7, 2018, 8:16 p.m. UTC
From: Junaid Shahid <junaids@google.com>

The per-cpu memcg stock can retain a charge of upto 32 pages. On a
machine with large number of cpus, this can amount to a decent amount
of memory. Additionally force_empty interface might be triggering
unneeded memcg reclaims.

Signed-off-by: Junaid Shahid <juanids@google.com>
Signed-off-by: Shakeel Butt <shakeelb@google.com>
---
 mm/memcontrol.c | 3 +++
 1 file changed, 3 insertions(+)

Comments

Shakeel Butt May 7, 2018, 8:19 p.m. UTC | #1
On Mon, May 7, 2018 at 1:16 PM Shakeel Butt <shakeelb@google.com> wrote:

> From: Junaid Shahid <junaids@google.com>

> The per-cpu memcg stock can retain a charge of upto 32 pages. On a
> machine with large number of cpus, this can amount to a decent amount
> of memory. Additionally force_empty interface might be triggering
> unneeded memcg reclaims.

> Signed-off-by: Junaid Shahid <juanids@google.com>

This should be "Junaid Shahid <junaids@google.com>".

> Signed-off-by: Shakeel Butt <shakeelb@google.com>
> ---
>   mm/memcontrol.c | 3 +++
>   1 file changed, 3 insertions(+)

> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index e2d33a37f971..2c3c69524b49 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -2841,6 +2841,9 @@ static int mem_cgroup_force_empty(struct mem_cgroup
*memcg)

>          /* we call try-to-free pages for make this cgroup empty */
>          lru_add_drain_all();
> +
> +       drain_all_stock(memcg);
> +
>          /* try to free all pages in this cgroup */
>          while (nr_retries && page_counter_read(&memcg->memory)) {
>                  int progress;
> --
> 2.17.0.441.gb46fe60e1d-goog
Michal Hocko May 9, 2018, 8:18 a.m. UTC | #2
On Mon 07-05-18 13:16:51, Shakeel Butt wrote:
> From: Junaid Shahid <junaids@google.com>
> 
> The per-cpu memcg stock can retain a charge of upto 32 pages. On a
> machine with large number of cpus, this can amount to a decent amount
> of memory. Additionally force_empty interface might be triggering
> unneeded memcg reclaims.
> 
> Signed-off-by: Junaid Shahid <juanids@google.com>
> Signed-off-by: Shakeel Butt <shakeelb@google.com>

Acked-by: Michal Hocko <mhocko@suse.com>

> ---
>  mm/memcontrol.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index e2d33a37f971..2c3c69524b49 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -2841,6 +2841,9 @@ static int mem_cgroup_force_empty(struct mem_cgroup *memcg)
>  
>  	/* we call try-to-free pages for make this cgroup empty */
>  	lru_add_drain_all();
> +
> +	drain_all_stock(memcg);
> +
>  	/* try to free all pages in this cgroup */
>  	while (nr_retries && page_counter_read(&memcg->memory)) {
>  		int progress;
> -- 
> 2.17.0.441.gb46fe60e1d-goog

Patch
diff mbox

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index e2d33a37f971..2c3c69524b49 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -2841,6 +2841,9 @@  static int mem_cgroup_force_empty(struct mem_cgroup *memcg)
 
 	/* we call try-to-free pages for make this cgroup empty */
 	lru_add_drain_all();
+
+	drain_all_stock(memcg);
+
 	/* try to free all pages in this cgroup */
 	while (nr_retries && page_counter_read(&memcg->memory)) {
 		int progress;