Re: [PATCH -mm v2 7/8] slub: make dead memcg caches discard free slabs immediately

[Joonsoo Kim]

On Fri, Jun 06, 2014 at 05:22:44PM +0400, Vladimir Davydov wrote:
> Since a dead memcg cache is destroyed only after the last slab allocated
> to it is freed, we must disable caching of empty slabs for such caches,
> otherwise they will be hanging around forever.
> 
> This patch makes SLUB discard dead memcg caches' slabs as soon as they
> become empty. To achieve that, it disables per cpu partial lists for
> dead caches (see put_cpu_partial) and forbids keeping empty slabs on per
> node partial lists by setting cache's min_partial to 0 on
> kmem_cache_shrink, which is always called on memcg offline (see
> memcg_unregister_all_caches).
> 
> Signed-off-by: Vladimir Davydov <vdavydov@xxxxxxxxxxxxx>
> Thanks-to: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
> ---
>  mm/slub.c |   20 ++++++++++++++++++++
>  1 file changed, 20 insertions(+)
> 
> diff --git a/mm/slub.c b/mm/slub.c
> index e46d6abe8a68..1dad7e2c586a 100644
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@ -2015,6 +2015,8 @@ static void unfreeze_partials(struct kmem_cache *s,
>  #endif
>  }
>  
> +static void flush_all(struct kmem_cache *s);
> +
>  /*
>   * Put a page that was just frozen (in __slab_free) into a partial page
>   * slot if available. This is done without interrupts disabled and without
> @@ -2064,6 +2066,21 @@ static void put_cpu_partial(struct kmem_cache *s, struct page *page, int drain)
>  
>  	} while (this_cpu_cmpxchg(s->cpu_slab->partial, oldpage, page)
>  								!= oldpage);
> +
> +	if (memcg_cache_dead(s)) {
> +               bool done = false;
> +               unsigned long flags;
> +
> +               local_irq_save(flags);
> +               if (this_cpu_read(s->cpu_slab->partial) == page) {
> +                       unfreeze_partials(s, this_cpu_ptr(s->cpu_slab));
> +                       done = true;
> +               }
> +               local_irq_restore(flags);
> +
> +               if (!done)
> +                       flush_all(s);
> +	}

Now, slab_free() is non-preemptable so flush_all() isn't needed.

Thanks.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/