On Wed, Jan 23 2013, Glauber Costa wrote:
> In order to maintain all the memcg bookkeeping, we need per-node
> descriptors, which will in turn contain a per-zone descriptor.
>
> Because we want to statically allocate those, this array ends up being
> very big. Part of the reason is that we allocate something large enough
> to hold MAX_NUMNODES, the compile time constant that holds the maximum
> number of nodes we would ever consider.
>
> However, we can do better in some cases if the firmware help us. This is
> true for modern x86 machines; coincidentally one of the architectures in
> which MAX_NUMNODES tends to be very big.
>
> By using the firmware-provided maximum number of nodes instead of
> MAX_NUMNODES, we can reduce the memory footprint of struct memcg
> considerably. In the extreme case in which we have only one node, this
> reduces the size of the structure from ~ 64k to ~2k. This is
> particularly important because it means that we will no longer resort to
> the vmalloc area for the struct memcg on defconfigs. We also have enough
> room for an extra node and still be outside vmalloc.
>
> One also has to keep in mind that with the industry's ability to fit
> more processors in a die as fast as the FED prints money, a nodes = 2
> configuration is already respectably big.
>
> [ v2: use size_t for size calculations ]
> Signed-off-by: Glauber Costa <glommer@xxxxxxxxxxxxx>
> Cc: Michal Hocko <mhocko@xxxxxxx>
> Cc: Kamezawa Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
> Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
> Cc: Greg Thelen <gthelen@xxxxxxxxxx>
> Cc: Hugh Dickins <hughd@xxxxxxxxxx>
> Cc: Ying Han <yinghan@xxxxxxxxxx>
> Cc: Mel Gorman <mgorman@xxxxxxx>
> Cc: Rik van Riel <riel@xxxxxxxxxx>
Reviewed-by: Greg Thelen <gthelen@xxxxxxxxxx>
> ---
> mm/memcontrol.c | 40 +++++++++++++++++++++++++---------------
> 1 file changed, 25 insertions(+), 15 deletions(-)
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 09255ec..09d8b02 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -172,7 +172,7 @@ struct mem_cgroup_per_node {
> };
>
> struct mem_cgroup_lru_info {
> - struct mem_cgroup_per_node *nodeinfo[MAX_NUMNODES];
> + struct mem_cgroup_per_node *nodeinfo[0];
It seems like a VM_BUG_ON() in mem_cgroup_zoneinfo() asserting that the
nid index is less than nr_node_ids would be good at catching illegal
indexes. I don't see any illegal indexes in your patch, but I fear that
someday a MAX_NUMNODES based for() loop might sneak in.
> };
>
> /*
> @@ -276,17 +276,6 @@ struct mem_cgroup {
> */
> struct res_counter kmem;
> /*
> - * Per cgroup active and inactive list, similar to the
> - * per zone LRU lists.
> - */
> - struct mem_cgroup_lru_info info;
> - int last_scanned_node;
> -#if MAX_NUMNODES > 1
> - nodemask_t scan_nodes;
> - atomic_t numainfo_events;
> - atomic_t numainfo_updating;
> -#endif
> - /*
> * Should the accounting and control be hierarchical, per subtree?
> */
> bool use_hierarchy;
> @@ -349,8 +338,29 @@ struct mem_cgroup {
> /* Index in the kmem_cache->memcg_params->memcg_caches array */
> int kmemcg_id;
> #endif
> +
> + int last_scanned_node;
> +#if MAX_NUMNODES > 1
> + nodemask_t scan_nodes;
> + atomic_t numainfo_events;
> + atomic_t numainfo_updating;
> +#endif
> + /*
> + * Per cgroup active and inactive list, similar to the
> + * per zone LRU lists.
> + *
> + * WARNING: This has to be the last element of the struct. Don't
> + * add new fields after this point.
> + */
> + struct mem_cgroup_lru_info info;
> };
>
> +static inline size_t memcg_size(void)
> +{
> + return sizeof(struct mem_cgroup) +
> + nr_node_ids * sizeof(struct mem_cgroup_per_node);
> +}
> +
Tangential question: why use inline here? I figure that modern
compilers are good at making inlining decisions.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>