linux.git/mm/memcontrol.c, branch v3.5

mm: fix kernel-doc warnings

2012-06-20T21:39:36+00:00

Fix kernel-doc warnings such as

  Warning(../mm/page_cgroup.c:432): No description found for parameter 'id'
  Warning(../mm/page_cgroup.c:432): Excess function parameter 'mem' description in 'swap_cgroup_record'

Signed-off-by: Wanpeng Li 
Cc: Randy Dunlap 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

memcg: fix use_hierarchy css_is_ancestor oops regression

2012-06-20T21:39:35+00:00

If use_hierarchy is set, reclaim testing soon oopses in css_is_ancestor()
called from __mem_cgroup_same_or_subtree() called from page_referenced():
when processes are exiting, it's easy for mm_match_cgroup() to pass along
a NULL memcg coming from a NULL mm->owner.

Check for that in __mem_cgroup_same_or_subtree().  Return true or false?
False because we cannot know if it was in the hierarchy, but also false
because it's better not to count a reference from an exiting process.

Signed-off-by: Hugh Dickins 
Acked-by: Johannes Weiner 
Acked-by: Konstantin Khlebnikov 
Acked-by: KAMEZAWA Hiroyuki 
Acked-by: Michal Hocko 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

memcg: decrement static keys at real destroy time

2012-05-29T23:22:28+00:00

We call the destroy function when a cgroup starts to be removed, such as
by a rmdir event.

However, because of our reference counters, some objects are still
inflight.  Right now, we are decrementing the static_keys at destroy()
time, meaning that if we get rid of the last static_key reference, some
objects will still have charges, but the code to properly uncharge them
won't be run.

This becomes a problem specially if it is ever enabled again, because now
new charges will be added to the staled charges making keeping it pretty
much impossible.

We just need to be careful with the static branch activation: since there
is no particular preferred order of their activation, we need to make sure
that we only start using it after all call sites are active.  This is
achieved by having a per-memcg flag that is only updated after
static_key_slow_inc() returns.  At this time, we are sure all sites are
active.

This is made per-memcg, not global, for a reason: it also has the effect
of making socket accounting more consistent.  The first memcg to be
limited will trigger static_key() activation, therefore, accounting.  But
all the others will then be accounted no matter what.  After this patch,
only limited memcgs will have its sockets accounted.

[akpm@linux-foundation.org: move enum sock_flag_bits into sock.h,
                            document enum sock_flag_bits,
                            convert memcg_proto_active() and memcg_proto_activated() to test_bit(),
                            redo tcp_update_limit() comment to 80 cols]
Signed-off-by: Glauber Costa 
Cc: Tejun Heo 
Cc: Li Zefan 
Acked-by: KAMEZAWA Hiroyuki 
Cc: Johannes Weiner 
Cc: Michal Hocko 
Acked-by: David Miller 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

memcg: always free struct memcg through schedule_work()

2012-05-29T23:22:28+00:00

Right now we free struct memcg with kfree right after a rcu grace period,
but defer it if we need to use vfree() to get rid of that memory area.  We
do that by need, because we need vfree to be called in a process context.

This patch unifies this behavior, by ensuring that even kfree will happen
in a separate thread.  The goal is to have a stable place to call the
upcoming jump label destruction function outside the realm of the
complicated and quite far-reaching cgroup lock (that can't be held when
holding either the cpu_hotplug.lock or jump_label_mutex)

[akpm@linux-foundation.org: tweak comment]
Signed-off-by: Glauber Costa 
Cc: Tejun Heo 
Cc: Li Zefan 
Acked-by: KAMEZAWA Hiroyuki 
Cc: Johannes Weiner 
Acked-by: Michal Hocko 
Cc: David Miller 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/memcg: apply add/del_page to lruvec

2012-05-29T23:22:28+00:00

Take lruvec further: pass it instead of zone to add_page_to_lru_list() and
del_page_from_lru_list(); and pagevec_lru_move_fn() pass lruvec down to
its target functions.

This cleanup eliminates a swathe of cruft in memcontrol.c, including
mem_cgroup_lru_add_list(), mem_cgroup_lru_del_list() and
mem_cgroup_lru_move_lists() - which never actually touched the lists.

In their place, mem_cgroup_page_lruvec() to decide the lruvec, previously
a side-effect of add, and mem_cgroup_update_lru_size() to maintain the
lru_size stats.

Whilst these are simplifications in their own right, the goal is to bring
the evaluation of lruvec next to the spin_locking of the lrus, in
preparation for a future patch.

Signed-off-by: Hugh Dickins 
Cc: KOSAKI Motohiro 
Acked-by: KAMEZAWA Hiroyuki 
Acked-by: Michal Hocko 
Acked-by: Konstantin Khlebnikov 
Cc: Johannes Weiner 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm/memcg: get_lru_size not get_lruvec_size

2012-05-29T23:22:28+00:00

Konstantin just introduced mem_cgroup_get_lruvec_size() and
get_lruvec_size(), I'm about to add mem_cgroup_update_lru_size(): but
we're dealing with the same thing, lru_size[lru].  We ought to agree on
the naming, and I do think lru_size is the more correct: so rename his
ones to get_lru_size().

Signed-off-by: Hugh Dickins 
Acked-by: KAMEZAWA Hiroyuki 
Acked-by: Konstantin Khlebnikov 
Acked-by: Michal Hocko 
Cc: KOSAKI Motohiro 
Cc: Johannes Weiner 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: memcg: print statistics from live counters

2012-05-29T23:22:28+00:00

Directly print statistics and event counters instead of going through an
intermediate accumulation stage into a separate array, which used to
require defining statistic items in more than one place.

[akpm@linux-foundation.org: checkpatch fixes]
Signed-off-by: Johannes Weiner 
Acked-by: Michal Hocko 
Acked-by: KAMEZAWA Hiroyuki 
Cc: Ying Han 
Cc: Tejun Heo 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: memcg: group swapped-out statistics counter logically

2012-05-29T23:22:28+00:00

The counter of currently swapped out pages in a memcg (hierarchy) is
sitting amidst ever-increasing event counters.  Move this item to the
other counters that reflect current state rather than history.

This technically breaks the kernel ABI, but hopefully nobody relies on the
order of items in memory.stat.

Signed-off-by: Johannes Weiner 
Acked-by: Michal Hocko 
Acked-by: KAMEZAWA Hiroyuki 
Cc: Ying Han 
Cc: Tejun Heo 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: memcg: keep ratelimit counter separate from event counters

2012-05-29T23:22:27+00:00

All events except the ratelimit counter are statistics exported to
userspace.  Keep this internal value out of the event count array.

Signed-off-by: Johannes Weiner 
Acked-by: Michal Hocko 
Acked-by: KAMEZAWA Hiroyuki 
Cc: Ying Han 
Cc: Tejun Heo 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: memcg: print statistics directly to seq_file

2012-05-29T23:22:27+00:00

Being able to use seq_printf() allows being smarter about statistics
name strings, which are currently listed twice, with the only difference
being a "total_" prefix on the hierarchical version.

Signed-off-by: Johannes Weiner 
Acked-by: Michal Hocko 
Acked-by: KAMEZAWA Hiroyuki 
Cc: Ying Han 
Cc: Tejun Heo 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds