On Tue 06-09-16 15:52:57, Vlastimil Babka wrote:
> The new ultimate compaction priority disables some heuristics, which may 
> result
> in excessive cost. This is fine for non-costly orders where we want to try 
> hard
> before resulting for OOM, but might be disruptive for costly orders which do
> not trigger OOM and should generally have some fallback. Thus, we disable the
> full priority for costly orders.
> 
> Suggested-by: Michal Hocko <mho...@kernel.org>
> Signed-off-by: Vlastimil Babka <vba...@suse.cz>
> Cc: Michal Hocko <mho...@kernel.org>
> Cc: Mel Gorman <mgor...@techsingularity.net>
> Cc: Joonsoo Kim <iamjoonsoo....@lge.com>
> Cc: David Rientjes <rient...@google.com>
> Cc: Rik van Riel <r...@redhat.com>

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

> ---
>  include/linux/compaction.h | 1 +
>  mm/page_alloc.c            | 5 ++++-
>  2 files changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/include/linux/compaction.h b/include/linux/compaction.h
> index 585d55cb0dc0..0d8415820fc3 100644
> --- a/include/linux/compaction.h
> +++ b/include/linux/compaction.h
> @@ -9,6 +9,7 @@ enum compact_priority {
>       COMPACT_PRIO_SYNC_FULL,
>       MIN_COMPACT_PRIORITY = COMPACT_PRIO_SYNC_FULL,
>       COMPACT_PRIO_SYNC_LIGHT,
> +     MIN_COMPACT_COSTLY_PRIORITY = COMPACT_PRIO_SYNC_LIGHT,
>       DEF_COMPACT_PRIORITY = COMPACT_PRIO_SYNC_LIGHT,
>       COMPACT_PRIO_ASYNC,
>       INIT_COMPACT_PRIORITY = COMPACT_PRIO_ASYNC
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index f8bed910e3cf..ff60a2837c58 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -3165,6 +3165,7 @@ should_compact_retry(struct alloc_context *ac, int 
> order, int alloc_flags,
>                    int compaction_retries)
>  {
>       int max_retries = MAX_COMPACT_RETRIES;
> +     int min_priority;
>  
>       if (!order)
>               return false;
> @@ -3204,7 +3205,9 @@ should_compact_retry(struct alloc_context *ac, int 
> order, int alloc_flags,
>        * if we exhausted all retries at the lower priorities
>        */
>  check_priority:
> -     if (*compact_priority > MIN_COMPACT_PRIORITY) {
> +     min_priority = (order > PAGE_ALLOC_COSTLY_ORDER) ?
> +                     MIN_COMPACT_COSTLY_PRIORITY : MIN_COMPACT_PRIORITY;
> +     if (*compact_priority > min_priority) {
>               (*compact_priority)--;
>               return true;
>       }
> -- 
> 2.9.3

-- 
Michal Hocko
SUSE Labs

Reply via email to