Re: [RFC PATCH 1/3] zram: charge the compressed RAM to the page's memcgroup

[Michal Hocko <mhocko@suse.com>]

On Thu 15-06-23 11:48:30, Zhongkun He wrote:
[...]
> @@ -1419,6 +1420,10 @@ static int zram_write_page(struct zram *zram, struct page *page, u32 index)
>  	struct zcomp_strm *zstrm;
>  	unsigned long element = 0;
>  	enum zram_pageflags flags = 0;
> +	struct mem_cgroup *memcg, *old_memcg;
> +
> +	memcg = page_memcg(page);
> +	old_memcg = set_active_memcg(memcg);
>  
>  	mem = kmap_atomic(page);
>  	if (page_same_filled(mem, &element)) {
[...]
> @@ -1470,8 +1475,10 @@ static int zram_write_page(struct zram *zram, struct page *page, u32 index)
>  		handle = zs_malloc(zram->mem_pool, comp_len,
>  				GFP_NOIO | __GFP_HIGHMEM |
>  				__GFP_MOVABLE);
> -		if (IS_ERR_VALUE(handle))
> -			return PTR_ERR((void *)handle);
> +		if (IS_ERR_VALUE(handle)) {
> +			ret = PTR_ERR((void *)handle);
> +			goto out;
> +		}
>  
>  		if (comp_len != PAGE_SIZE)
>  			goto compress_again;

I am not really deeply familiar with zram implementation nor usage but
how is the above allocation going to be charged without __GFP_ACCOUNT in
the gfp mask?

Also what exactly is going to happen for the swap backed by the zram
device? Your memcg might be hitting the hard limit and therefore
swapping out. Wouldn't zs_malloc fail very likely under that condition
making the swap effectively unusable?
-- 
Michal Hocko
SUSE Labs