Commit 4bdfc1c4a943ce8707675ad510ea1076c9e8e528

Authored by Vladimir Davydov
Committed by Linus Torvalds
1 parent 24d404dc10

memcg: fix destination cgroup leak on task charges migration

We are supposed to take one css reference per each memory page and per
each swap entry accounted to a memory cgroup.  However, during task
charges migration we take a reference to the destination cgroup twice
per each swap entry: first in mem_cgroup_do_precharge()->try_charge()
and then in mem_cgroup_move_swap_account(), permanently leaking the
destination cgroup.

The hunk taking the second reference seems to be a leftover from the
pre-00501b531c472 ("mm: memcontrol: rewrite charge API") era.  Remove it
to fix the leak.

Fixes: e8ea14cc6ead (mm: memcontrol: take a css reference for each charged page)
Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Showing 1 changed file with 0 additions and 12 deletions Side-by-side Diff

... ... @@ -3043,18 +3043,6 @@
3043 3043 if (swap_cgroup_cmpxchg(entry, old_id, new_id) == old_id) {
3044 3044 mem_cgroup_swap_statistics(from, false);
3045 3045 mem_cgroup_swap_statistics(to, true);
3046   - /*
3047   - * This function is only called from task migration context now.
3048   - * It postpones page_counter and refcount handling till the end
3049   - * of task migration(mem_cgroup_clear_mc()) for performance
3050   - * improvement. But we cannot postpone css_get(to) because if
3051   - * the process that has been moved to @to does swap-in, the
3052   - * refcount of @to might be decreased to 0.
3053   - *
3054   - * We are in attach() phase, so the cgroup is guaranteed to be
3055   - * alive, so we can just call css_get().
3056   - */
3057   - css_get(&to->css);
3058 3046 return 0;
3059 3047 }
3060 3048 return -EINVAL;