2023-06-21 08:12:36

by liuq

Subject: [PATCH v2] mm/min_free_kbytes: modify min_free_kbytes calculation rules

The current calculation of min_free_kbytes only uses ZONE_DMA and
ZONE_NORMAL pages,but the ZONE_MOVABLE zone->_watermark[WMARK_MIN]
will also divide part of min_free_kbytes.This will cause the min
watermark of ZONE_NORMAL to be too small in the presence of ZONE_MOVEABLE.

__GFP_HIGH and PF_MEMALLOC allocations usually don't need moveable
zone pages, so just like ZONE_HIGHMEM, cap pages_min to a small
value in __setup_per_zone_wmarks.

Signed-off-by: liuq <[email protected]>
---
mm/page_alloc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 47421bedc12b..608384712a89 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -6364,7 +6364,7 @@ static void __setup_per_zone_wmarks(void)

/* Calculate total number of !ZONE_HIGHMEM pages */
for_each_zone(zone) {
- if (!is_highmem(zone))
+ if (!is_highmem(zone) || zone_idx(zone) != ZONE_MOVABLE)
lowmem_pages += zone_managed_pages(zone);
}

@@ -6374,7 +6374,7 @@ static void __setup_per_zone_wmarks(void)
spin_lock_irqsave(&zone->lock, flags);
tmp = (u64)pages_min * zone_managed_pages(zone);
do_div(tmp, lowmem_pages);
- if (is_highmem(zone)) {
+ if (is_highmem(zone) || zone_idx(zone) == ZONE_MOVABLE) {
/*
* __GFP_HIGH and PF_MEMALLOC allocations usually don't
* need highmem pages, so cap pages_min to a small
--
2.27.0

2023-06-21 08:30:45

by Huang, Ying

Subject: Re: [PATCH v2] mm/min_free_kbytes: modify min_free_kbytes calculation rules

Hi, Liuq,

Thanks for updated patch.

liuq <[email protected]> writes:

> The current calculation of min_free_kbytes only uses ZONE_DMA and
> ZONE_NORMAL pages,but the ZONE_MOVABLE zone->_watermark[WMARK_MIN]
> will also divide part of min_free_kbytes.This will cause the min
> watermark of ZONE_NORMAL to be too small in the presence of ZONE_MOVEABLE.
>
> __GFP_HIGH and PF_MEMALLOC allocations usually don't need moveable
> zone pages, so just like ZONE_HIGHMEM, cap pages_min to a small
> value in __setup_per_zone_wmarks.

Please list the test result of your patch. That is, WMARK_MIN before
and after your patch in a system with large ZONE_MOVABLE.

> Signed-off-by: liuq <[email protected]>
> ---
> mm/page_alloc.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 47421bedc12b..608384712a89 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -6364,7 +6364,7 @@ static void __setup_per_zone_wmarks(void)
>
> /* Calculate total number of !ZONE_HIGHMEM pages */
> for_each_zone(zone) {
> - if (!is_highmem(zone))
> + if (!is_highmem(zone) || zone_idx(zone) != ZONE_MOVABLE)
> lowmem_pages += zone_managed_pages(zone);
> }
>
> @@ -6374,7 +6374,7 @@ static void __setup_per_zone_wmarks(void)
> spin_lock_irqsave(&zone->lock, flags);
> tmp = (u64)pages_min * zone_managed_pages(zone);
> do_div(tmp, lowmem_pages);
> - if (is_highmem(zone)) {
> + if (is_highmem(zone) || zone_idx(zone) == ZONE_MOVABLE) {
> /*
> * __GFP_HIGH and PF_MEMALLOC allocations usually don't
> * need highmem pages, so cap pages_min to a small
~~~~~~~

Change the comments above too?

Best Regards,
Huang, Ying

2023-06-21 08:40:11

by liuq

Subject: Re: [PATCH v2] mm/min_free_kbytes: modify min_free_kbytes calculation rules

在 2023/6/21 16:09, liuq 写道:
> Hi, Liuq,
>
> Thanks for updated patch.
>
> liuq <[email protected]> writes:
>
> > The current calculation of min_free_kbytes only uses ZONE_DMA and
> > ZONE_NORMAL pages,but the ZONE_MOVABLE zone->_watermark[WMARK_MIN]
> > will also divide part of min_free_kbytes.This will cause the min
> > watermark of ZONE_NORMAL to be too small in the presence of
> ZONE_MOVEABLE.
> >
> > __GFP_HIGH and PF_MEMALLOC allocations usually don't need moveable
> > zone pages, so just like ZONE_HIGHMEM, cap pages_min to a small
> > value in __setup_per_zone_wmarks.
>
> Please list the test result of your patch.  That is, WMARK_MIN before
> and after your patch in a system with large ZONE_MOVABLE.
>
Hi Huang, Ying

Thank you very much for your review and suggestions. I will send another
version later.

Best Regards,

liuq

> > Signed-off-by: liuq <[email protected]>
> > ---
> >  mm/page_alloc.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> > index 47421bedc12b..608384712a89 100644
> > --- a/mm/page_alloc.c
> > +++ b/mm/page_alloc.c
> > @@ -6364,7 +6364,7 @@ static void __setup_per_zone_wmarks(void)
> >
> >          /* Calculate total number of !ZONE_HIGHMEM pages */
> >          for_each_zone(zone) {
> > -                if (!is_highmem(zone))
> > +                if (!is_highmem(zone) || zone_idx(zone) !=
> ZONE_MOVABLE)
> >                          lowmem_pages += zone_managed_pages(zone);
> >          }
> >
> > @@ -6374,7 +6374,7 @@ static void __setup_per_zone_wmarks(void)
> >                  spin_lock_irqsave(&zone->lock, flags);
> >                  tmp = (u64)pages_min * zone_managed_pages(zone);
> >                  do_div(tmp, lowmem_pages);
> > -                if (is_highmem(zone)) {
> > +                if (is_highmem(zone) || zone_idx(zone) ==
> ZONE_MOVABLE) {
> >                          /*
> >                           * __GFP_HIGH and PF_MEMALLOC allocations
> usually don't
> >                           * need highmem pages, so cap pages_min to
> a small
>                                 ~~~~~~~
>
> Change the comments above too?
>
> Best Regards,
> Huang, Ying
>

2023-06-21 11:29:07

by David Hildenbrand

Subject: Re: [PATCH v2] mm/min_free_kbytes: modify min_free_kbytes calculation rules

On 21.06.23 09:39, liuq wrote:
> The current calculation of min_free_kbytes only uses ZONE_DMA and
> ZONE_NORMAL pages,but the ZONE_MOVABLE zone->_watermark[WMARK_MIN]
> will also divide part of min_free_kbytes.This will cause the min
> watermark of ZONE_NORMAL to be too small in the presence of ZONE_MOVEABLE.
>
> __GFP_HIGH and PF_MEMALLOC allocations usually don't need moveable
> zone pages, so just like ZONE_HIGHMEM, cap pages_min to a small
> value in __setup_per_zone_wmarks.
>
> Signed-off-by: liuq <[email protected]>
> ---
> mm/page_alloc.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 47421bedc12b..608384712a89 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -6364,7 +6364,7 @@ static void __setup_per_zone_wmarks(void)
>
> /* Calculate total number of !ZONE_HIGHMEM pages */
> for_each_zone(zone) {
> - if (!is_highmem(zone))
> + if (!is_highmem(zone) || zone_idx(zone) != ZONE_MOVABLE)
> lowmem_pages += zone_managed_pages(zone);
> }
>
> @@ -6374,7 +6374,7 @@ static void __setup_per_zone_wmarks(void)
> spin_lock_irqsave(&zone->lock, flags);
> tmp = (u64)pages_min * zone_managed_pages(zone);
> do_div(tmp, lowmem_pages);
> - if (is_highmem(zone)) {
> + if (is_highmem(zone) || zone_idx(zone) == ZONE_MOVABLE) {
> /*
> * __GFP_HIGH and PF_MEMALLOC allocations usually don't
> * need highmem pages, so cap pages_min to a small

Patch subject should probably be something like

"mm/page_alloc: fix min_free_kbytes calculation regarding ZONE_MOVABLE"

?

--
Cheers,

David / dhildenb