On Fri, Mar 18, 2011 at 10:29:50AM -0400, Vivek Goyal wrote:
> On Thu, Mar 17, 2011 at 01:43:50PM +0100, Johannes Weiner wrote:
>
> [..]
> > So structures roughly like this:
> >
> > struct mem_cgroup {
> > ...
> > /* key is struct backing_dev_info * */
> > struct rb_root memcg_bdis;
> > };
> >
> > struct memcg_bdi {
> > /* key is struct address_space * */
> > struct rb_root memcg_mappings;
> > struct rb_node node;
> > };
> >
> > struct memcg_mapping {
> > struct address_space *mapping;
> > struct mem_cgroup *memcg;
> > struct rb_node node;
> > atomic_t count;
> > };
> >
> > struct page_cgroup {
> > ...
> > struct memcg_mapping *memcg_mapping;
> > };
>
> Johannes, didn't you want page->mapping to point to memcg_mapping instead
> of increasing the size of page_cgroup?
Initially, yes, but this is far less invasive. We don't increase
page_cgroup, though: memcg_mapping contains the pointer to struct
mem_cgroup, it can replace pc->memcg for now.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>