lkml.org 
[lkml]   [2020]   [Sep]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH 6/9] mm, page_alloc: cache pageset high and batch in struct zone
From
Date
On 22.09.20 16:37, Vlastimil Babka wrote:
> All per-cpu pagesets for a zone use the same high and batch values, that are
> duplicated there just for performance (locality) reasons. This patch adds the
> same variables also to struct zone as a shared copy.
>
> This will be useful later for making possible to disable pcplists temporarily
> by setting high value to 0, while remembering the values for restoring them
> later. But we can also immediately benefit from not updating pagesets of all
> possible cpus in case the newly recalculated values (after sysctl change or
> memory online/offline) are actually unchanged from the previous ones.
>
> Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
> ---
> include/linux/mmzone.h | 6 ++++++
> mm/page_alloc.c | 16 ++++++++++++++--
> 2 files changed, 20 insertions(+), 2 deletions(-)
>
> diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
> index 90721f3156bc..7ad3f14dbe88 100644
> --- a/include/linux/mmzone.h
> +++ b/include/linux/mmzone.h
> @@ -470,6 +470,12 @@ struct zone {
> #endif
> struct pglist_data *zone_pgdat;
> struct per_cpu_pageset __percpu *pageset;
> + /*
> + * the high and batch values are copied to individual pagesets for
> + * faster access
> + */
> + int pageset_high;
> + int pageset_batch;
>
> #ifndef CONFIG_SPARSEMEM
> /*
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index de3b48bda45c..901907799bdc 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -5824,6 +5824,8 @@ static void build_zonelists(pg_data_t *pgdat)
> * Other parts of the kernel may not check if the zone is available.
> */
> static void pageset_init(struct per_cpu_pageset *p);
> +#define BOOT_PAGESET_HIGH 0
> +#define BOOT_PAGESET_BATCH 1

Much better. A comment would have been nice ("this disables the pcp via
the boot pageset completely.") :) (I'm pretty sure I'd forget at one
point what these values mean)

> static DEFINE_PER_CPU(struct per_cpu_pageset, boot_pageset);
> static DEFINE_PER_CPU(struct per_cpu_nodestat, boot_nodestats);
>
> @@ -6213,8 +6215,8 @@ static void pageset_init(struct per_cpu_pageset *p)
> * need to be as careful as pageset_update() as nobody can access the
> * pageset yet.
> */
> - pcp->high = 0;
> - pcp->batch = 1;
> + pcp->high = BOOT_PAGESET_HIGH;
> + pcp->batch = BOOT_PAGESET_BATCH;
> }
>
> /*
> @@ -6238,6 +6240,14 @@ static void zone_set_pageset_high_and_batch(struct zone *zone)
> new_batch = max(1UL, 1 * new_batch);
> }
>
> + if (zone->pageset_high != new_high ||
> + zone->pageset_batch != new_batch) {
> + zone->pageset_high = new_high;
> + zone->pageset_batch = new_batch;
> + } else {
> + return;
> + }
> +
> for_each_possible_cpu(cpu) {
> p = per_cpu_ptr(zone->pageset, cpu);
> pageset_update(&p->pcp, new_high, new_batch);
> @@ -6300,6 +6310,8 @@ static __meminit void zone_pcp_init(struct zone *zone)
> * offset of a (static) per cpu variable into the per cpu area.
> */
> zone->pageset = &boot_pageset;
> + zone->pageset_high = BOOT_PAGESET_HIGH;
> + zone->pageset_batch = BOOT_PAGESET_BATCH;

I do wonder if copying from any cpuvar inside boot_pageset is cleaner.

zone->pageset_high = &this_cpu_ptr(zone->pageset)->pcp.high;
...


--
Thanks,

David / dhildenb

\
 
 \ /
  Last update: 2020-09-25 12:34    [W:1.720 / U:0.288 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site