[Devel] Re: [PATCH 05/11] Add a __GFP_KMEMCG flag

[David Rientjes]

On Mon, 25 Jun 2012, Glauber Costa wrote:

> This flag is used to indicate to the callees that this allocation will be
> serviced to the kernel. It is not supposed to be passed by the callers
> of kmem_cache_alloc, but rather by the cache core itself.
> 

Not sure what "serviced to the kernel" means, does this mean that the 
memory will not be accounted for to the root memcg?

> CC: Christoph Lameter <cl at linux.com>
> CC: Pekka Enberg <penberg at cs.helsinki.fi>
> CC: Michal Hocko <mhocko at suse.cz>
> CC: Kamezawa Hiroyuki <kamezawa.hiroyu at jp.fujitsu.com>
> CC: Johannes Weiner <hannes at cmpxchg.org>
> CC: Suleiman Souhlal <suleiman at google.com>
> ---
>  include/linux/gfp.h |    8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/include/linux/gfp.h b/include/linux/gfp.h
> index 1e49be4..8f4079f 100644
> --- a/include/linux/gfp.h
> +++ b/include/linux/gfp.h
> @@ -37,6 +37,9 @@ struct vm_area_struct;
>  #define ___GFP_NO_KSWAPD	0x400000u
>  #define ___GFP_OTHER_NODE	0x800000u
>  #define ___GFP_WRITE		0x1000000u
> +#ifdef CONFIG_CGROUP_MEM_RES_CTLR_KMEM
> +#define ___GFP_KMEMCG		0x2000000u
> +#endif
>  
>  /*
>   * GFP bitmasks..
> @@ -88,13 +91,16 @@ struct vm_area_struct;
>  #define __GFP_OTHER_NODE ((__force gfp_t)___GFP_OTHER_NODE) /* On behalf of other node */
>  #define __GFP_WRITE	((__force gfp_t)___GFP_WRITE)	/* Allocator intends to dirty page */
>  
> +#ifdef CONFIG_CGROUP_MEM_RES_CTLR_KMEM
> +#define __GFP_KMEMCG	((__force gfp_t)___GFP_KMEMCG)/* Allocation comes from a memcg-accounted resource */
> +#endif

Needs a space.

>  /*
>   * This may seem redundant, but it's a way of annotating false positives vs.
>   * allocations that simply cannot be supported (e.g. page tables).
>   */
>  #define __GFP_NOTRACK_FALSE_POSITIVE (__GFP_NOTRACK)
>  
> -#define __GFP_BITS_SHIFT 25	/* Room for N __GFP_FOO bits */
> +#define __GFP_BITS_SHIFT 26	/* Room for N __GFP_FOO bits */
>  #define __GFP_BITS_MASK ((__force gfp_t)((1 << __GFP_BITS_SHIFT) - 1))
>  
>  /* This equals 0, but use constants in case they ever change */