linux-kernel - Re: [PATCH v4 3/5] zsmalloc: Add a LRU to zs_pool to keep track of zspages in LRU order

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Y3ayb4kx/m3oDsSN@google.com>
Date:   Thu, 17 Nov 2022 14:15:11 -0800
From:   Minchan Kim <minchan@...nel.org>
To:     Nhat Pham <nphamcs@...il.com>
Cc:     akpm@...ux-foundation.org, hannes@...xchg.org, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org, ngupta@...are.org,
        senozhatsky@...omium.org, sjenning@...hat.com, ddstreet@...e.org,
        vitaly.wool@...sulko.com
Subject: Re: [PATCH v4 3/5] zsmalloc: Add a LRU to zs_pool to keep track of
 zspages in LRU order

On Thu, Nov 17, 2022 at 08:38:37AM -0800, Nhat Pham wrote:
> This helps determines the coldest zspages as candidates for writeback.
> 
> Signed-off-by: Nhat Pham <nphamcs@...il.com>
> ---
>  mm/zsmalloc.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 46 insertions(+)
> 
> diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
> index 326faa751f0a..2557b55ec767 100644
> --- a/mm/zsmalloc.c
> +++ b/mm/zsmalloc.c
> @@ -239,6 +239,11 @@ struct zs_pool {
>  	/* Compact classes */
>  	struct shrinker shrinker;
> 
> +#ifdef CONFIG_ZPOOL
> +	/* List tracking the zspages in LRU order by most recently added object */
> +	struct list_head lru;
> +#endif
> +
>  #ifdef CONFIG_ZSMALLOC_STAT
>  	struct dentry *stat_dentry;
>  #endif
> @@ -260,6 +265,12 @@ struct zspage {
>  	unsigned int freeobj;
>  	struct page *first_page;
>  	struct list_head list; /* fullness list */
> +
> +#ifdef CONFIG_ZPOOL
> +	/* links the zspage to the lru list in the pool */
> +	struct list_head lru;
> +#endif
> +
>  	struct zs_pool *pool;
>  #ifdef CONFIG_COMPACTION
>  	rwlock_t lock;
> @@ -352,6 +363,18 @@ static void cache_free_zspage(struct zs_pool *pool, struct zspage *zspage)
>  	kmem_cache_free(pool->zspage_cachep, zspage);
>  }
> 
> +#ifdef CONFIG_ZPOOL
> +/* Moves the zspage to the front of the zspool's LRU */
> +static void move_to_front(struct zs_pool *pool, struct zspage *zspage)
> +{
> +	assert_spin_locked(&pool->lock);
> +
> +	if (!list_empty(&zspage->lru))
> +		list_del(&zspage->lru);
> +	list_add(&zspage->lru, &pool->lru);
> +}
> +#endif
> +
>  /* pool->lock(which owns the handle) synchronizes races */
>  static void record_obj(unsigned long handle, unsigned long obj)
>  {
> @@ -953,6 +976,9 @@ static void free_zspage(struct zs_pool *pool, struct size_class *class,
>  	}
> 
>  	remove_zspage(class, zspage, ZS_EMPTY);
> +#ifdef CONFIG_ZPOOL
> +	list_del(&zspage->lru);
> +#endif
>  	__free_zspage(pool, class, zspage);
>  }
> 
> @@ -998,6 +1024,10 @@ static void init_zspage(struct size_class *class, struct zspage *zspage)
>  		off %= PAGE_SIZE;
>  	}
> 
> +#ifdef CONFIG_ZPOOL
> +	INIT_LIST_HEAD(&zspage->lru);
> +#endif
> +
>  	set_freeobj(zspage, 0);
>  }
> 
> @@ -1418,6 +1448,11 @@ unsigned long zs_malloc(struct zs_pool *pool, size_t size, gfp_t gfp)
>  		fix_fullness_group(class, zspage);
>  		record_obj(handle, obj);
>  		class_stat_inc(class, OBJ_USED, 1);
> +
> +#ifdef CONFIG_ZPOOL
> +		/* Move the zspage to front of pool's LRU */
> +		move_to_front(pool, zspage);
> +#endif
>  		spin_unlock(&pool->lock);
> 
>  		return handle;
> @@ -1444,6 +1479,10 @@ unsigned long zs_malloc(struct zs_pool *pool, size_t size, gfp_t gfp)
> 
>  	/* We completely set up zspage so mark them as movable */
>  	SetZsPageMovable(pool, zspage);
> +#ifdef CONFIG_ZPOOL
> +	/* Move the zspage to front of pool's LRU */
> +	move_to_front(pool, zspage);
> +#endif
>  	spin_unlock(&pool->lock);

Why do we move the zspage in the alloc instead of accessor?

Isn't zs_map_object better place since it's clear semantic
that user start to access the object?

If you are concerning unnecessary churning, can we do that
only for WO access?

Yeah, it is still weird layering since allocator couldn't know
what user will do over the this access(will keep or discard)
so the lru in the allocator is not a good design but I just
want to make little more sense.