Re: [PATCH v1 2/3] mm/memory_hotplug: don't shuffle complete zone when onlining memory

[Michal Hocko]

On Tue 16-06-20 13:52:12, David Hildenbrand wrote:
> Commit e900a918b098 ("mm: shuffle initial free memory to improve
> memory-side-cache utilization") introduced shuffling of free pages
> during system boot and whenever we online memory blocks.
> 
> However, whenever we online memory blocks, all pages that will be
> exposed to the buddy end up getting freed via __free_one_page(). In the
> general case, we free these pages in MAX_ORDER - 1 chunks, which
> corresponds to the shuffle order.
> 
> Inside __free_one_page(), we will already shuffle the newly onlined pages
> using "to_tail = shuffle_pick_tail();". Drop explicit zone shuffling on
> memory hotplug.
> 
> Note: When hotplugging a DIMM, each memory block (e.g., 128MB .. 2G on
> x86-64) will get onlined individually, resulting in a shuffle_zone() for
> every memory block getting onlined.
> 
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Cc: Alexander Duyck <alexander.h.duyck@xxxxxxxxxxxxxxx>
> Cc: Dan Williams <dan.j.williams@xxxxxxxxx>
> Cc: Michal Hocko <mhocko@xxxxxxxx>
> Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
> Cc: Kees Cook <keescook@xxxxxxxxxxxx>
> Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
> Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>

Acked-by: Michal Hocko <mhocko@xxxxxxxx>

> ---
>  mm/memory_hotplug.c |  3 ---
>  mm/shuffle.c        |  2 +-
>  mm/shuffle.h        | 12 ------------
>  3 files changed, 1 insertion(+), 16 deletions(-)
> 
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index 9b34e03e730a4..845a517649c71 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -40,7 +40,6 @@
>  #include <asm/tlbflush.h>
>  
>  #include "internal.h"
> -#include "shuffle.h"
>  
>  /*
>   * online_page_callback contains pointer to current page onlining function.
> @@ -822,8 +821,6 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages,
>  	zone->zone_pgdat->node_present_pages += onlined_pages;
>  	pgdat_resize_unlock(zone->zone_pgdat, &flags);
>  
> -	shuffle_zone(zone);
> -
>  	node_states_set_node(nid, &arg);
>  	if (need_zonelists_rebuild)
>  		build_all_zonelists(NULL);
> diff --git a/mm/shuffle.c b/mm/shuffle.c
> index dd13ab851b3ee..609c26aa57db0 100644
> --- a/mm/shuffle.c
> +++ b/mm/shuffle.c
> @@ -180,7 +180,7 @@ void __meminit __shuffle_free_memory(pg_data_t *pgdat)
>  	struct zone *z;
>  
>  	for (z = pgdat->node_zones; z < pgdat->node_zones + MAX_NR_ZONES; z++)
> -		shuffle_zone(z);
> +		__shuffle_zone(z);
>  }
>  
>  bool shuffle_pick_tail(void)
> diff --git a/mm/shuffle.h b/mm/shuffle.h
> index 4d79f03b6658f..657e2b9ec38dd 100644
> --- a/mm/shuffle.h
> +++ b/mm/shuffle.h
> @@ -30,14 +30,6 @@ static inline void shuffle_free_memory(pg_data_t *pgdat)
>  	__shuffle_free_memory(pgdat);
>  }
>  
> -extern void __shuffle_zone(struct zone *z);
> -static inline void shuffle_zone(struct zone *z)
> -{
> -	if (!static_branch_unlikely(&page_alloc_shuffle_key))
> -		return;
> -	__shuffle_zone(z);
> -}
> -
>  static inline bool is_shuffle_order(int order)
>  {
>  	if (!static_branch_unlikely(&page_alloc_shuffle_key))
> @@ -54,10 +46,6 @@ static inline void shuffle_free_memory(pg_data_t *pgdat)
>  {
>  }
>  
> -static inline void shuffle_zone(struct zone *z)
> -{
> -}
> -
>  static inline void page_alloc_shuffle(enum mm_shuffle_ctl ctl)
>  {
>  }
> -- 
> 2.26.2

-- 
Michal Hocko
SUSE Labs