Prev: [PATCH 1/7] memcg: sc.nr_to_reclaim should be initialized
Next: memcg: nid and zid can be calculated from zone
From: KAMEZAWA Hiroyuki on 16 Jul 2010 06:30
On Fri, 16 Jul 2010 19:13:31 +0900 (JST)
KOSAKI Motohiro <kosaki.motohiro(a)jp.fujitsu.com> wrote:

> Currently, mem_cgroup_shrink_node_zone() initialize sc.nr_to_reclaim as 0.
> It mean shrink_zone() only scan 32 pages and immediately return even if
> it doesn't reclaim any pages.
>
> This patch fixes it.
>
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro(a)jp.fujitsu.com>

Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu(a)jp.fujitsu.com>

> ---
> mm/vmscan.c | 1 +
> 1 files changed, 1 insertions(+), 0 deletions(-)
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 1691ad0..bd1d035 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -1932,6 +1932,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
> struct zone *zone, int nid)
> {
> struct scan_control sc = {
> + .nr_to_reclaim = SWAP_CLUSTER_MAX,
> .may_writepage = !laptop_mode,
> .may_unmap = 1,
> .may_swap = !noswap,
> --
> 1.6.5.2
>
>
>
>

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Mel Gorman on 16 Jul 2010 06:30
On Fri, Jul 16, 2010 at 07:13:31PM +0900, KOSAKI Motohiro wrote:
> Currently, mem_cgroup_shrink_node_zone() initialize sc.nr_to_reclaim as 0.
> It mean shrink_zone() only scan 32 pages and immediately return even if
> it doesn't reclaim any pages.
>

Do you mean it immediately returns once one page is reclaimed? i.e. this
check

if (nr_reclaimed >= nr_to_reclaim && priority < DEF_PRIORITY)
break;


> This patch fixes it.
>

Otherwise it seems ok. It's unrelated to trace points though so should
be submitted on its own.

> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro(a)jp.fujitsu.com>

Acked-by: Mel Gorman <mel(a)csn.ul.ie>

> ---
> mm/vmscan.c | 1 +
> 1 files changed, 1 insertions(+), 0 deletions(-)
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 1691ad0..bd1d035 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -1932,6 +1932,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
> struct zone *zone, int nid)
> {
> struct scan_control sc = {
> + .nr_to_reclaim = SWAP_CLUSTER_MAX,
> .may_writepage = !laptop_mode,
> .may_unmap = 1,
> .may_swap = !noswap,
> --
> 1.6.5.2
>
>
>

--
Mel Gorman
Part-time Phd Student Linux Technology Center
University of Limerick IBM Dublin Software Lab
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: KOSAKI Motohiro on 21 Jul 2010 09:40
> On Fri, Jul 16, 2010 at 07:13:31PM +0900, KOSAKI Motohiro wrote:
> > Currently, mem_cgroup_shrink_node_zone() initialize sc.nr_to_reclaim as 0.
> > It mean shrink_zone() only scan 32 pages and immediately return even if
> > it doesn't reclaim any pages.
> >
>
> Do you mean it immediately returns once one page is reclaimed? i.e. this
> check
>
> if (nr_reclaimed >= nr_to_reclaim && priority < DEF_PRIORITY)
> break;

Strictly speaking, once SWAP_CLUSTER_MAX batch is scanned. no need
to reclaim pages at all.



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Balbir Singh on 22 Jul 2010 01:40
* KOSAKI Motohiro <kosaki.motohiro(a)jp.fujitsu.com> [2010-07-16 19:13:31]:

> Currently, mem_cgroup_shrink_node_zone() initialize sc.nr_to_reclaim as 0.
> It mean shrink_zone() only scan 32 pages and immediately return even if
> it doesn't reclaim any pages.
>
> This patch fixes it.
>
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro(a)jp.fujitsu.com>
> ---
> mm/vmscan.c | 1 +
> 1 files changed, 1 insertions(+), 0 deletions(-)
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 1691ad0..bd1d035 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -1932,6 +1932,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
> struct zone *zone, int nid)
> {
> struct scan_control sc = {
> + .nr_to_reclaim = SWAP_CLUSTER_MAX,
> .may_writepage = !laptop_mode,
> .may_unmap = 1,
> .may_swap = !noswap,

Could you please do some additional testing on

1. How far does this push pages (in terms of when limit is hit)?
2. Did you hit a problem with the current setting or is it a review
fix?


--
Three Cheers,
Balbir
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: KOSAKI Motohiro on 23 Jul 2010 03:40
> * KOSAKI Motohiro <kosaki.motohiro(a)jp.fujitsu.com> [2010-07-16 19:13:31]:
>
> > Currently, mem_cgroup_shrink_node_zone() initialize sc.nr_to_reclaim as 0.
> > It mean shrink_zone() only scan 32 pages and immediately return even if
> > it doesn't reclaim any pages.
> >
> > This patch fixes it.
> >
> > Signed-off-by: KOSAKI Motohiro <kosaki.motohiro(a)jp.fujitsu.com>
> > ---
> > mm/vmscan.c | 1 +
> > 1 files changed, 1 insertions(+), 0 deletions(-)
> >
> > diff --git a/mm/vmscan.c b/mm/vmscan.c
> > index 1691ad0..bd1d035 100644
> > --- a/mm/vmscan.c
> > +++ b/mm/vmscan.c
> > @@ -1932,6 +1932,7 @@ unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
> > struct zone *zone, int nid)
> > {
> > struct scan_control sc = {
> > + .nr_to_reclaim = SWAP_CLUSTER_MAX,
> > .may_writepage = !laptop_mode,
> > .may_unmap = 1,
> > .may_swap = !noswap,
>
> Could you please do some additional testing on
>
> 1. How far does this push pages (in terms of when limit is hit)?

32 pages per mem_cgroup_shrink_node_zone().

That said, the algorithm is here.

1. call mem_cgroup_largest_soft_limit_node()
calculate largest cgroup
2. call mem_cgroup_shrink_node_zone() and shrink 32 pages
3. goto 1 if limit is still exceed.

If it's not your intention, can you please your intended algorithm?


> 2. Did you hit a problem with the current setting or is it a review
> fix?

I've found this by review. and my patch works fine on my test environment.
Of cource, if you do _not_ run the code on heavy pressure, your original code
works too.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 |  Next  |  Last
Pages: 1 2
Prev: [PATCH 1/7] memcg: sc.nr_to_reclaim should be initialized
Next: memcg: nid and zid can be calculated from zone